View
1
Download
0
Category
Preview:
Citation preview
The influence of place of
death and rural residency on
the ‘good death’
Suzanne Rainsford
July 2017
A thesis submitted for the degree of Doctor of Philosophy of
The Australian National University
©Suzanne Rainsford 2017
All rights reserved
i
Statement of authenticity
I, Suzanne Rainsford, declare that this thesis, submitted in fulfilment of the
requirements for the award of Doctor of Philosophy, Medical School, College of
Medicine Biology and Environment, The Australian University, is the product of my
own work. All assistance received in preparing this thesis has been acknowledged
and/or referenced. This thesis has not been submitted for qualifications at any other
academic institution.
Signature of PhD candidate________________________________________________
ii
iii
Acknowledgements
It is with immense gratitude I thank the many people who assisted me along the thesis
journey. Without their support, encouragement and willingness to participate,
completion of this thesis would not have been possible.
I am forever grateful to Emeritus Professor, Nicholas Glasgow, chair of my supervisory
panel, who encouraged me from the first day I mentioned the crazy idea of undertaking
a PhD. At the very beginning, he gave me an honest picture of what the journey would
look like, and has been supportive through the highs and lows of research life. I am also
sincerely appreciative of my supervisors, Associate Professor Christine Phillips,
Professor Rod MacLeod, and Dr Robert Wiles. You have all been so encouraging, and
generous with your time and knowledge. I will always remember your eagerness to be
part of the supervisory panel. Never once did you hesitate in participating in yet another
systematic literature review. Without your ongoing support this thesis would never have
become a reality.
Other members of the academic community were also generous in sharing their time
and expertise. In the early days of this study, Dr Graham Fordham, ANU Medical
School, graciously permitted me to attend his Qualitative Methodologies in Health
Research tutorials. Dr Terry Neeman, ANU Statistical Consulting Unit, patiently guided
me through the maze of statistical analysis. Professor Donna Wilson, Faculty of
Nursing, University of Alberta, Canada, graciously accepted the invitation to be an
external advisor - her insight and commentary were much appreciated.
I am also appreciative of the support provided by Cathy Burton (ANU Library), Lidia
Hrvatin (Librarian, Calvary HealthCare Bruce), Danielle (who found transcribing to be
a “touching” experience), Sally Josh (who provided assistance with the Local Health
District HREC application process), and Cheryl Morse (Executive Support, Office of
the Dean, who enthusiastically arranged my supervisor appointments with Nick).
This research study would not have been possible without the participation of the 29
Snowy Monaro residents, who were either living, or caring for someone, with a life-
limiting illness. They were the true heroes. Despite their illness, distress, fatigue and
iv
burden, they willingly shared their stories, experiences and perspectives. It is my hope
that what they have shared will play a part in improving the rural end-of-life care for
those who will sadly follow in their footsteps. I am also appreciative of the support of
the local funeral directors, general practitioners, nurses, and residential aged care
managers.
To my family and friends, I apologise for neglecting you over the past three years. I
hope you forgive me and I look forward to re-establishing rich fellowship again. I have
been aware of your love, patience, and quiet support, encouragement and understanding.
Finally, I wish to thank the Australian National University, and all the unidentified
people who made this incredible opportunity possible. It has certainly been an enriching
and fulfilling experience.
This research was supported by an Australian Government Research Training Program
(RTP) Scholarship.
v
Abstract
Background
Achieving one’s preferred place of death is often considered, within palliative care, as a
proxy for a ‘good death’. To date, most place of death studies are quantitative and
reflect an urban view. The objectives of this study were to explore the concept of the
good death from the experiences and perspectives of rural patients with a life limiting
illness, and their family caregivers, within the Snowy Monaro region of New South
Wales, Australia; to determine the influence of place of death and rural residency on the
‘good death’.
Methods
This thesis was a mixed methods study, comprising
1. Two systematic and one scoping literature reviews, exploring rural place of
death, the rural ‘good death’, and end-of-life care from the experiences and
perspectives of rural patients and family caregivers.
2. An ethnographic study utilising open ended interviews with 11 rural patients
with life limiting illness, 18 family caregivers, and six clinicians (medical and
nursing).
3. A cross-sectional study of cause and place of death for all people who died in
the Snowy Monaro region between 1st February 2015 and 31st May 2016
(n = 224).
Results
Of the 224 deaths recorded, 138 (62%) were due to an illness amenable to palliative
care. Seventeen (12%) of these deaths occurred at home, in a private residence; 52
(38%) occurred in the usual place of residence. Residential aged care was considered
home for some residents. Home was the initial preferred place of death for those
interviewed. Over time, dying in a safe place became more important than dying at
home. A ‘safe death’ emerged as the central theme of a ‘good death’.
Home is more than the material structure. It represents connection to the land, family,
and memories, all elements that underpin its value as a safe place. Rural residency
helped maintain home as a safe place (e.g. privacy on the farm); however, these same
features quickly rendered home an unsafe place of death (e.g. isolation). The roles of the
vi
rural hospital and residential aged care in end-of-life care are unique, and their
familiarity within one’s community often makes them a safer alternative to home, and
substitutes for in-patient hospice.
Conclusion
Most Snowy Monaro residents do not die at home, yet a ‘good death’ is achievable.
Many elements of the ‘good death’ are not unique to rural residency; however rural
perspectives on place of death are often contrary to the urban view. A death at home
does not ensure a good death. The task for all those providing and supporting end-of-life
care is to ensure all places for dying can deliver the ‘safe death’ no matter where the
illness trajectory dictates that the person dies.
vii
Publications arising from this research
1. Rainsford, S., MacLeod, R.D., & Glasgow, N.J. (2016). Place of death in rural
palliative care: a systematic review. Palliative Medicine, 30(8), 745-63.
doi:10.1177/0269216316628779.
2. Rainsford, S., MacLeod, R.D., Glasgow, N.J., Wilson, D.M., Phillips, C.B., &
Wiles, R.B. (2016). Rural residents' perspectives on the rural 'good death': a
scoping review. Health and Social Care in the Community.
doi: 10.1111/hsc.12385. [Epub ahead of print, 19 September 2016].
3. Rainsford, S., MacLeod, R.D., Glasgow, N.J., Phillips, C.B., Wiles, R.B., &
Wilson, D.M. (2017). Rural end-of-life care from the experiences and
perspectives of patients and family caregivers: a systematic literature review.
Palliative Medicine; 31(10): 895-912. doi: 10.1177/0269216316685234.
Conference oral presentations
1. Rainsford, S., MacLeod, R.D., Glasgow, N.J. (2016). Does dying at home matter
most? 22nd Hospice New Zealand Palliative Care Conference. 14-16 September
2016. Auckland, NZ.
Conference posters
1. Rainsford, S., Macleod, R.D., Glasgow, N.J. (2016). Place of death in a rural
setting. Australian and New Zealand Society of Palliative Medicine Conference,
8 – 11 September 2016, Perth, Australia.
2. Rainsford, S., Phillips, C., Glasgow, N., MacLeod, R., Wiles, R., Wilson, DM.
(2017). Place of death does not determine the ‘good death’ for rural family
caregivers. An ethnographic study. 15th World Congress of the European
Association for Palliative Care, 18 - 20 May 2017, Madrid, Spain.
viii
3. Rainsford, S., Glasgow, N., MacLeod, R., Phillips, C., Neeman, T., Wiles, R.
(2017). “I want to stay home and die on the farm.” The actual place of death in
a rural setting. 15th World Congress of the European Association for Palliative
Care, 18 - 20 May 2017, Madrid, Spain.
Author contribution to publications
The systematic and scoping literature reviews were my original idea. I developed the
research questions for each publication. I conducted the systematic reviews; developed
the search terms and inclusion criteria; conducted the database searches, review of titles,
abstracts, and full texts, quality assessment, charting, thematic analysis, and back
calculations (Publication 1). I wrote the peer reviewed articles, and responded to the
reviewers’ comments.
NJ Glasgow (Chair of supervisory panel): co-authored all papers and conference
presentations arising from the thesis. Independently reviewed some identified abstracts;
independently verified data extraction and quality assessment; reviewed manuscript
drafts prior to submission.
RD MacLeod (Supervisor): co-authored all papers and conference presentations arising
from the thesis. Independently reviewed all identified titles and some abstracts;
independently verified data extraction and quality assessment; reviewed manuscript
drafts prior to submission.
CB Phillips and RB Wiles (Supervisors): Independently reviewed and assessed content,
data extraction and quality of a portion of identified abstracts. Reviewed co-authored
manuscript drafts prior to submission.
DM Wilson (External advisor): Offered advice on conducting a scoping review;
independently reviewed and assessed content and quality of a portion of identified
abstracts. Reviewed co-authored manuscript drafts prior to submission.
T Neeman (ANU Statistical Consultancy Unit): Provided advice on quantitative data
analysis and reporting.
ix
CONTENTS
STATEMENT OF AUTHENTICITY i
ACKNOWLEDGEMENTS iii
ABSTRACT v
PUBLICATIONS ARISING FROM THESIS vii
TABLE OF CONTENTS ix
LIST OF APPENDICES xiii
LIST OF TABLES xiv
LIST OF FIGURES xv
LIST OF ABBREVIATIONS xvi
PART 1: INTRODUCTION TO RESEARCH STUDY 1
Chapter One: Introduction 3
Introduction 3
1.1 Researcher’s world view 4
1.2 Research objectives 9
1.3 Research method 9
1.3.1 Literature review 9
1.3.2 Study setting 10
1.3.3 Ethnography 10
1.3.4 Population death data survey 11
1.4 Challenges specific to this study 12
1.5 Thesis structure 12
PART 2: RESEARCH PREPARATION,
METHODOLOGY AND METHODS 17
Introduction to Part 2 19
Chapter Two: Literature review 21
Introduction 21
Chapter structure 22
2.1 Palliative Care 22
2.1.1 Origins of palliative care 23
2.1.2 Contemporary palliative care 24
2.1.3 Advance care planning 26
2.1.4 The quality of palliative care 27
2.2 The ‘good death’ 28
2.2.1 Evolution of the concept of the good death 28
2.2.2 Contemporary ‘good death’ 30
2.3 Challenges in defining rural 31
2.4 Rural residents’ perspectives on the rural ‘good death’: a scoping
review 34
2.5 Place of death in rural palliative care: a systematic review 37
x
2.6 Rural end-of-life care from the perspectives of patients and family
caregivers: a systematic review 44
2.6.1 Patients’ experiences and perspectives 47
2.6.1.1 Maintaining Dignity 47
2.6.1.2 Preparing for death 48
2.6.1.3 Pain control 48
2.6.1.4 Formal and informal support 48
2.6.2 Family caregivers’ experiences and perspectives 48
2.6.3 Patient and family caregivers’ common experiences and
perspectives 49
2.6.3.1 Communication, accurate and timely
information 49
2.6.3.2 Formal services 50
2.6.3.3 Informal social support - family,
friends, and neighbours 51
2.6.3.4 Emotional support 51
2.6.3.5 Spiritual support 52
2.6.4 Influence of rural place and culture 52
2.6.5 Meaning of home or ‘home country’ 53
2.6.6 Summary of rural end-of-life care from the
experiences of patients and family caregivers 53
2.7 Australian rural studies 54
2.8 Chapter summary 54
Chapter Three: The Snowy Monaro – describing the setting
and population 57
3.1 Introduction 57
3.2 The Snowy Monaro region 58
3.2.1 Snowy Monaro history 60
3.2.2 Medical facilities 64
3.2.3 Transport 68
3.2.4 Community services 69
3.2.5 Snowy Monaro region summary 69
3.3 The target population 69
3.4 Challenges in recruiting palliative care patients to research 70
3.5 Chapter summary 72
Chapter Four: Research methodology 73
Introduction 73
4.1 Research objectives and questions 74
4.2 Systematic and scoping reviews 74
4.2.1. The rural ‘good death’: a scoping review 75
4.2.1.1 Eligibility criteria 76
4.2.1.2 Information sources and search strategy 77
4.2.1.3 Assessment of data quality 78
4.2.1.4 Synthesis 79
xi
4.2.2. Systematic reviews 79
4.2.2.1. Place of death in rural palliative care: a systematic
review 79
4.2.2.2. Rural end-of-life care from the experiences and
perspectives of patients and family caregivers: a
systematic review 82
4.3. Theoretical framework and methodology of the mixed methods
research study 85
4.3.1 Ethnography 86
4.4 Ethnographic component 88
4.4.1 Recruitment 88
4.4.2 Participant consent 91
4.4.3 Interview participants 93
4.4.4 Data collection - the interviews 93
4.4.5 Analysis of qualitative data 96
4.5 Quantitative component 97
4.5.1 Quantitative methods objectives 98
4.5.2 Recruitment 98
4.5.3 Quantitative data collection 99
4.5.4 Quantitative analysis 100
4.6 Ethical considerations 102
4.7 The role of the researcher 102
4.8 Limitations 104
4.9 Rigour of the research 104
PART 3: RESEARCH STUDY RESULTS 107
Introduction to Part 3 109
Chapter Five: Place of death for Snowy Monaro residents 111
Introduction 111
5.1 Description of all deaths 111
5.2 Non-palliative subgroup 113
5.3 Palliative subgroup 114
5.3.1 Place of death according to cause of death 115
5.3.2 Place of death according to degree of rurality 115
5.3.3 Deaths in usual place of residence 118
5.3.4 Factors influencing death at usual place of residence 118
5.3.5 Factors influencing death within region 119
5.4 Deaths in residential aged care facilities 120
5.5 Palliative deaths, excluding permanent RACF residents 121
5.6 Limitations 123
5.7 Chapter summary 124
Chapter Six: Place of death and the meaning of place 127
Gwen’s story 127
Introduction 129
xii
6.1. Actual place of death 129
6.2. Preferred place of death 130
6.3. The meaning of home 134
6.4. Hospital as the place of death 141
6.5. Residential aged care as the place of death 149
6.6. Inpatient hospice as the place of death 154
6.7. Congruence between actual and preferred place of death 155
6.8. Concordance between preferences of patients, family caregivers,
and healthcare professionals 156
6.9. Chapter summary 157
Chapter Seven: The rural ‘good death’ 159
Ivan’s story 159
Introduction 161
7.1. What is a ‘good death’? 161
7.2. Factors associated with a ‘good death’ 164
7.2.1. Symptom control 165
7.2.2. Place of death 169
7.2.3. Family 172
7.2.4. Community support 176
7.2.5. Autonomy 178
7.2.6. Spirituality 179
7.3. The central theme of the ‘good death’ 181
7.4. Chapter summary 182
Chapter Eight: The influence of rural residency on the ‘good
death’ and place of death 185
Marianne’s story 185
Introduction 187
8.1. Patients’ place of residence 187
8.2. Rural identity and resilience 188
8.3. Challenges of rural residency 189
8.3.1. Distance and travel 190
8.3.2. Other challenges of rural residency 197
8.3.2.1. Lack of local specialist services 197
8.3.2.2. Lack of local medical expertise 197
8.3.2.3. Lack of allied health support 198
8.4. Benefits of rural residency 199
8.4.1. Personalised care 200
8.4.2. Informal community support 202
8.4.3. Formal support 204
8.4.4. Convenience of rural residency 206
8.5. Rural residency and safety 207
8.6. Chapter summary 210
xiii
PART 4: DISCUSSION AND IMPLICATIONS 211
Chapter Nine: Discussion and implications 213
Introduction 213
9.1. Place of death for Snowy Monaro residents 214
9.2. Meaning of place 218
9.3. The rural ‘good death’ 224
9.4. Influence of rural residency on the ‘good death’ and place
of death 225
9.5. Limitations 226
9.6. Reflexivity 227
9.7. Implications for policy and practice 230
REFERENCES 233
LIST OF APPENDICES 253
Appendix 1: HREC Ethics approvals 254
Appendix 2: Quality assessment score cards for the systematic literature
reviews 261
Appendix 3: Numerical summary of the studies included in the
‘good death’ scoping review 263
Appendix 4: Participant information sheets and consent form 264
Appendix 5: Interview prompts 275
Appendix 6: Copies of the published papers 277
xiv
LIST OF TABLES
Table 2.1: Studies included in the ‘good death’ scoping literature review 35
Table 2.2: Studies included in the actual PoD systematic literature review 39
Table 2.3: Studies included in the preferred PoD systematic literature review 41
Table 2.4: Studies included in the rural end-of-life systematic literature
review 45
Table 3.1. Estimated resident population (ERP) by remoteness structure 58
Table 3.2. Demographic features of Snowy Monaro residents 60
Table 4.1. Inclusion criteria for interviews 90
Table 4.2. List of participants and demographic features of the patients
(interviewed and identified in obituary notices) 94
Table 4.3. Informants of the Snowy Monaro death data 100
Table 5.1. Decedents’ demographics and place of death 112
Table 5.2. Place of death for the palliative subgroup by socio-demographic
characteristics 116
Table 5.3. Odds ratio estimates and 95% confidence limits of factors
associated with dying in usual place of residence and in region, for
the palliative subgroup 119
Table 5.4. Characteristics and PoD for all permanent RACF residents 120
Table 5.5. Socio-demographics of palliative sub-groups: including, excluding,
and exclusive of permanent residential aged care facility residents 122
xv
LIST OF FIGURES Figure 3.1. Map of NSW with study location indicated by arrow 59
Figure 3.2. Map of the Snowy Monaro Region 60
Figure 3.3. Monaro Plains, outside Cooma (2016) 61
Figure 3.4. The Main Range, Kosciuszko National Park 61
Figure 3.5. Cattle drive on the Snowy Mountains Highway 63
Figure 3.6. Cooma; population 6742 (August 2016) 64
Figure 3.7. Delegate; population 351 (August 2016) 66
Figure 3.8. Sharp Street Surgery, Cooma and Perisher Valley Medical
Centre 67
Figure 3.9. Great Diving Range, view from outskirts of Cooma 70
Figure 3.10. Monaro Dorper lambs 72
Figure 4.1. Flow diagram of scoping review selection process including
reasons for exclusion 78
Figure 4.2. Selection flow chart for place of death systematic literature
review 82
Figure 4.3. Selection flow chart for rural EoL systematic literature
review 84
Figure 4. 4. Theoretical framework and methods underpinning the
complete research study 86
Figure 4.5. Recruitment flow chart for interviews 89
Figure 4.6. Coding tree 97
Figure 4.7. SPSS modelling for logistic regression 102
Figure 5.1. Cause of death of the non-palliative subgroup 113
Figure 5.2. Deaths caused by a condition amenable to palliative care 114
Figure 5.3. Deaths in usual place of residence according to cause of
death for the palliative subgroup 118
Figure 7.1. Interaction of themes and the creation of the safe ‘good
death’ 182
xvi
Abbreviations
ABS Australian Bureau of Statistics
ACI Agency for Clinical Innovation
ACP Advance care planning
ANZSPM Australian and New Zealand Society of Palliative Medicine
APoD Actual place of death
CCF Congestive cardiac failure
CDH Cooma District Hospital
CHH Clare Holland House (out of region inpatient hospice)
CN Community nurse
CNC Clinical nurse consultant
COPD Chronic obstructive pulmonary disease
CVA Cerebral vascular accident
DON Director of nursing
EoL End-of-life
EoLC End-of-life care
ERP Estimated resident population
ESRF End stage renal failure
FAChPM Fellow of the Australasian Chapter of Palliative Medicine
FCG Family caregivers (family member or person identified by patient
as primary informal caregiver)
FTE Full-time equivalent
GP General practitioner
GWNSW LHD Greater Western New South Wales Local Health District
HREC Human research ethics committee
ICU Intensive care unit
HCP Health care professional
LGA Local government area
LHD Local health district
LOS Length of stay
MCCD Medical Certificate Cause of Death
MCCR Monaro Committee for Cancer Research.
MND Motor neuron disease
xvii
MPS Multipurpose service
MRHS Monaro Regional Health Services
NSW New South Wales
PCN Palliative care nurse
PoC Place of care
PoD Place of death
PPoC Preferred place of care
PPoD Preferred place of death
PRISMA Preferred Reporting Items for Systematic Reviews and Meta-
Analysis
QODD Quality of death and dying
RAC Residential aged care
RACF Residential aged care facility
SAH Subarachnoid haemorrhage
SNSW LHD Southern NSW Local Health District
TCH The Canberra Hospital
UPoR Usual place of residence
WHO World Health Organisation
1
PART 1: RESEARCH INTRODUCTION
2
3
CHAPTER ONE:
INTRODUCTION TO RESEARCH STUDY
‘Approaches to death and dying reveal much of the attitude of society as a
whole, to the individuals who compose it’. (Dame Cicely Saunders, foreword to
Oxford textbook of Palliative Medicine. Doyle et al. 1999)
Introduction
The World Health Organisation (WHO) defines Palliative Care as ‘an approach that
improves the quality of life of patients and their families facing the problems associated
with life-threatening illness’ (WHO, 2004). One goal of palliative care is to enable
people to die in their preferred place, with the literature suggesting home is the
preferred place for most (Gomes et al., 2013). However, most people don’t die at home,
with most deaths in the developed world occurring in institutions, such as hospitals and
residential aged care facilities (Broad et al., 2013; Cohen et al., 2015). Some even
suggest that a hospital death is suboptimal (Earle et al., 2003), and imply that a ‘good
death’ is a death that occurs at home.
Having lived and worked in the rural Snowy Monaro region of New South Wales
(NSW), Australia, for the past 25 years, I have observed that most of my dying patients
and acquaintances, died in the local district hospital. While a few remained home, until
death, most of these terminally ill residents had remained home until symptoms, or
family inability to cope, rendered home inappropriate. It was taken for granted they
would transfer to the local hospital to die. Over the years, from my personal and
professional perspective, the local rural hospital has been a very appropriate and
supportive place to die. However, recently in Australia, and in many developed
countries, a discourse has emerged in the popular media and in the professional
literature about encouraging home deaths by increasing the support to help people die at
home, respecting patients’ choices, and reducing the number of expected, often
expensive, deaths in hospitals. As I pondered this dilemma, I wondered why the rural
4
hospital had seemed appropriate for the people I had encountered, and if rural patients
and their family caregivers (FCGs) have a different perspective on hospital deaths and
what it means to ‘die well’. And so, the basis of my thesis was conceived. The principal
questions to be answered through this research study are: (1) what is a rural ‘good
death’? and (2) what is the influence of place of death and rural residency on the ‘good
death’?
This introductory chapter firstly outlines my personal world view on place of death
(PoD) (Section 1.1), as it is important from the outset that I declare any researcher bias I
may bring to this research study. This is followed by a statement of the research study
objectives (Section 1.2). Section 1.3 provides an overview of the research methodology.
The chapter concludes (Section 1.4) with a brief overview of the successive chapters
contained within this thesis.
1.1 Researcher world view
I grew up in the city. I don’t remember my first introduction to death, as I was a toddler
when my last living grandparent died. Grandma lived with us, adding to the business of
a household with five children. Eventually, our wise general practitioner (GP) suggested
that respite in a nursing home should be considered, so Dad and Mum could take a well-
deserved family holiday, a break from the exhausting daily responsibilities before the
arrival of their sixth child. Grandma never returned home; she died a few weeks later. I
have a sense Mum always regretted, even felt guilty for, not having cared for her mother
at home in the last days of her life.
After graduating from medical school, in the mid-1980s, I spent the then obligatory two
years as a junior hospital doctor, before stepping out into general practice. My interest
lay in sports medicine, caring for young, healthy, and fit patients. After six months
working at an elite sports medicine clinic, I eagerly headed off to the NSW Snowy
Mountains (Australia), to become a rural GP, working in the ski fields. I have vague
memories of caring for a few patients with cancer. In those days, there was no formal
palliative care, nor was there much discussion with patients as to their preferences re
place of care or death. If patients had stable symptoms and merely ‘faded away’ they
remained at home. Alternatively, and more commonly, they were admitted to the district
5
hospital, over 60 kilometres away. Elderly patients, who could no longer care for
themselves at home, were admitted to the local residential aged care facility (RACF).
Life in the ski fields was challenging, but fun. However, 1997 was, to quote Queen
Elizabeth II, my ‘annus horribilis’. The year started with the tragic death of my teenage
niece in a car accident. This was followed a few weeks later with the destruction, by
fire, of the house of elderly friends. A landslide in the ski village of Thredbo, in which
several of my patients died, followed a few months later. Despite the amazing
community spirit and camaraderie in such tragic circumstances, suddenly the gloss of
working in the mountains was lost. An opening became available to work on a tropical
island, over two thousand kilometres away, just off the coast of Townsville in far north
Queensland.
It was during my time on the island that I was exposed to the difficulties of providing
palliative care in an isolated area. We weren’t really that isolated, just a 20-minute ferry
ride to the mainland, between the hours of 6am and 9pm. I was the only GP caring for a
population of 3000, eight kilometres off shore. There was a public hospital outpatient
clinic staffed by a nurse and resident on-call doctor, as well as one ambulance officer.
After-hours emergencies could be transferred to the mainland by police launch,
providing the waters were calm.
I don’t remember the exact details, as I have done well to block out the trauma of
looking after Joe, a 65-year old man (names and details have been altered). He lived
with his wife, Jane who was in her early 60s. His regular GP was on the mainland. Joe
had recently been diagnosed with pancreatic cancer and had spent a few weeks in
Townsville hospital, for pain management. One afternoon I received a frantic phone call
from Jane, asking me to visit. Joe was in terrible pain. The first time I met Joe he was
thrashing about in bed and vomiting. He was unable to take his oral pain medication. I
finally settled him with an injection, most likely pethidine and metoclopramide, which
was standard practice at the time. Joe had told me he was not going back to hospital,
where he was just a patient in a large public ward. He wanted to remain home, on the
island. I made a few calls to the newly established pain management clinic who had
cared for Joe during his admission. The impression was that Joe was a ‘difficult case’
and his pain had not been adequately controlled. A new medication was suggested, but
6
really “you’re on your own…good luck.” The community nurse looked to me for
guidance, but nothing seemed to help Joe. At times, all we could do was ‘just be there’
for Joe and Jane. About 4 weeks after first meeting Joe, I received a call from Jane at
about 2am. Joe was in a lot of pain, could I come quickly? Joe died 10 minutes later,
just before I arrived. At that point, I didn’t want anything more to do with palliative
care. I had failed Joe and Jane.
I returned to the Snowy Mountains and joined a general practice in Cooma, about 100
kilometres from the ski fields. I managed to avoid palliative care. I had not applied for
accreditation at the 30-bed district hospital. I lost contact with patients once they
became ‘palliative’. Once admitted to hospital, ongoing care was provided by the
admitting GP. Then one day Beth came to see me.
Beth was 40 years old. She had grown up in Cooma, and until recently had been living
and working in the city. She had been diagnosed a few years earlier with cancer.
Unfortunately, the cancer had metastasised. Beth had brought her husband and young
family back home, to spend her last few months with her mother and brother. Beth was
complicated…she had fistulas, multiple ‘ostomies and stents. She was not absorbing her
opiate tablets. Although a challenge, caring for Beth was different to Joe. Beth wanted
to stay home, no matter what she faced, and we succeeded in respecting those wishes.
Beth died at home a few months later, surrounded by her family. This time there was
support - a great palliative care community nurse, a wonderful palliative medicine
specialist on the end of the phone in Sydney, GP colleagues in the practice, and an array
of drug options. This ignited my desire to pursue palliative care. Despite Beth’s family
and the palliative care nurse saying I “did well”, I knew I could have done things better.
In between these two significant patients I also had the opportunity to witness palliative
care in a more personal way. My elderly dad had cancer, and died suddenly during one
of his frequent admissions to a large private Sydney hospital. He had been cared for,
both at home and in hospital, by a palliative medicine specialist. Dad did not want to die
at home. To him the hospital was a familiar and safe place. Dad didn’t want home
‘turned into a hospital’ with beds and equipment. He didn’t want Mum to have the full
responsibility of caring for him during his frequent pain episodes. He had one respite
7
admission in the nearby hospice, but this was not a safe place for him. Sharing a four-
bedded room meant Dad and Mum grieved each time a ‘room-mate’ died.
In 2012, 25 years after graduation, I started palliative care specialist training. It was
always my intention to return to my rural town once training was completed. In mid-
2014, I returned to Cooma and started a private, part-time palliative medicine practice,
offering consultations in the surgery, private homes, hospital and RACFs.
Much is talked about the ‘good death’ and preferred place of death (PoD). From my
perspective, Joe did not have a ‘good’ death, despite being at home, in his preferred
place. From my perspective, Beth’s death was not ‘good’, things could have been done
better. Her pain could possibly have been better managed in hospital. That is not the
perception of the family or the palliative care nurse. For them, it was difficult because
symptoms were not easily controlled, but ‘good’ because she died at home. They had
felt extremely supported by their palliative care nurse and doctor. In the end, Beth had a
‘good enough death’ (McNamara, 2004). The family were satisfied and appreciative of
the effort that was made to keep Beth at home. My dad died in a private tertiary hospital
- that was the right place for him; much better there than at home or on a four- bed ward
at the hospice. During my palliative care training, I looked after Jimmy. He was 26
years old, and died in a side room, not hooked up to life support machinery, in the
intensive care unit (ICU) in an acute tertiary hospital. My palliative care team said his
death in ICU was not ‘good’, and that we had failed him. But, it was the right place for
someone young, who had an acute onset of respiratory distress, who wanted to fight
until the end, until all possible active treatments had been exhausted.
From my perspective, formulated through personal and professional experience, the
PoD should suite the circumstances, and be the most appropriate place at the time, for
the patient and their family. Place of death is dependent on each person’s needs and
coping ability. While it is imperative to respect a person’s autonomy and wishes, as
health care professionals (HCPs) we need to not only listen to our patients, but also
offer information, advice, and guidance to assist them in making good decisions. I
personally question the value of enquiring of one’s preferred PoD too far ahead of time,
because it is impossible to know what the scenarios will be. Asking people for their
preferences risks giving the impression their preference will be honoured. If the
8
preferred place is not achievable, or available, there is often a sense of disappointment,
or even guilt held by the family. I’ve seen families struggle trying to keep a promise
made to a patient; I’ve seen patients struggle not wanting to disappoint their family by
leaving home and going to a hospital or hospice. I’ve seen and experienced HCPs’
disappointment and a knowing better care could have been provided in an alternative
place. I’ve seen ‘good’ and ‘bad’ deaths at home, in hospital, in RACFs, and even in the
inpatient hospice.
Readers of this thesis may ask why I have indulged in writing about my experiences. I
have chosen to undertake an ethnographic study, and this methodology is open to
researcher bias. It is therefore important I declare my history and my position, to be
transparent.
From my perspective, issues surrounding patients and FCGs’ preferences are not the
only contentious areas. National and international studies report the actual PoD often
from death certificate data; however, this only provides the actual PoD, not the reasons
for determining the place. The data show the predictors for PoD, but again, not the
reasons. The data do not indicate whether decedents were cared for at home, only to
transfer to hospital for the final hours or days before death. Also, many people have
seen or heard of a family or friend dying at home, peacefully in their sleep. If they relied
on that one experience, they may falsely believe a home death, in any situation, is
uncomplicated and peaceful. Healthy population surveys are potentially meaningless if
the respondents have no personal experience of inadequately controlled symptoms, or
carer burden, fatigue, sacrifice, and responsibility often associated with a home death.
As previously stated, this research study was partly undertaken to answer my own
questions. As a rural palliative medicine specialist, I want to ensure my practice is
evidenced based and patient focused. My patients are rural residents. Do they, as the
literature suggests, want to die at home? Do they, as the literature suggests, believe the
small, rural community hospital is an inappropriate place to die? Am I, and my local
HCP colleagues failing our patients by not facilitating more deaths at home? Is it
possible for rural residents with a terminal illness to achieve a ‘good death’, regardless
of where they die? If so many Snowy Monaro residents are dying in the district
hospitals or local RACFs, does more need to be done to improve the quality of the
9
dying experience within the institutions, for the patient, their FCGs, and the staff? As I
delved deeper into the literature it became apparent that not only were there limited
studies reporting from the experiences and perspectives of patients and their FCGs, but
also the rural voices were significantly under-represented. As I pondered my questions,
the objectives of this research study were formulated.
1.2 Research objectives
The primary objective of this research study was to explore the concept of the good
death from the experiences and perspectives of rural patients with a life limiting illness,
and their FCGs, within the rural Snowy Monaro region of NSW, Australia. The
secondary objectives were to determine the influence of PoD and rural residency on the
‘good death’, and to explore the meaning behind the PoD preferences and decision-
making process. These perspectives were contextualised by describing the actual PoD of
Snowy Monaro residents, and by exploring the factors associated with PoD.
The following section provides an outline of the approach utilised to conduct this
research study and to address the research objectives.
1.3 Research methods
This thesis used a mixed methods approach. The research study was informed by a
number of systematic reviews of the literature. The actual PoD of Snowy Monaro
residents was described after analysing population death survey data, over a 16-month
period. It was important to have participants in this study describe their experiences in
their own way, and so a qualitative component was utilised based on ethnography.
Ethics approval was obtained from two independent human research ethics committees
(HREC).
1.3.1 Literature review
Two systematic and one scoping literature reviews were undertaken, using the Preferred
Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) and PRISMA-
Scoping Review guidelines (Equator Network, 2016). These literature reviews, now
published in peer-reviewed journals, (1) Place of death in rural palliative care: a
10
systematic review (Rainsford, MacLeod and Glasgow, 2016); (2) Rural end-of-life care
from the experiences and perspectives of patients and family caregivers: a systematic
literature review (Rainsford et al., 2017); and (3) Rural residents’ perspectives on the
rural ‘good death’: a scoping review (Rainsford et al., 2016), highlighted the paucity of
rural end-of-life studies, and the lack of findings reporting the experiences and
perspectives of rural patients, with a life limiting illness, and their FCGs. They also
illustrated the challenge of finding a single rural voice, as currently there is no
international consensus on the definition of ‘rural’, with significant heterogeneity across
rural studies. The identified rural studies, consistent with urban studies, reported that
most people with a life limiting illness do not die at home. However, there were
inconsistencies, with most studies reporting rural residents were less likely to die in
hospital than urban residents, while a few studies reported the opposite. The literature
suggests the ‘good death’ is subjective, and dependent on individual perspectives and
priorities. Dying in one’s preferred PoD is one feature of a ‘good death’; however, there
is insufficient evidence to determine the absolute importance of PoD in facilitating a
‘good death’. Most rural locations do not have a dedicated inpatient hospice unit, with
rural hospitals often acting as a substitute for inpatient hospice, albeit without a
specialist multidisciplinary palliative care team.
The scoping and systematic literature reviews identified gaps within the current rural
literature, worthy of further research. The research setting and methodologies are
outlined in the following sections, with detailed descriptions provided in Chapters Three
and Four.
1.3.2 Study setting
The setting chosen for the study was the Snowy Monaro region, located in south eastern
NSW, Australia. This covers an area of 15,162 km2 with a population of 20,218 (ABS,
2017a). It is serviced by the Monaro Regional Health Service. Cooma is the regional
centre and largest town, with a population of 6742 (ABS, 2017a). There are several
other smaller towns and communities, and many farms.
1.3.3 Ethnography
A qualitative approach, utilising open-ended interviews as the prime method of data
collection, was chosen to ensure sufficient depth and detail to fully understand each
11
participants’ personal experience (Patton 2002, pp. 16-17). Wilson et al. (2009a) report
that rural residents consider themselves to hold unique perspectives on death and dying,
and end-of-life care (EoLC). Therefore, this research study explores the influence of
rural culture and PoD on the concept of the good death, achieved through an
ethnographic approach.
Data were collected in 42 interviews. While the focus was to gain the perspectives of
patients and FCGs, local HCPs (palliative and community nurses, and GPs) were also
interviewed. In total, 35 participants (11 patients with a life expectancy of less than six
month, 15 of their FCGs, three additional bereaved FCGs, and six local HCPs) were
interviewed. Potential patient participants were identified by the local palliative care
nurse or GPs. After participants provided informed consent, they were interviewed
using a semi-structured interview format. Over time, recruitment became more
purposeful, to ensure the participants represented a cross-section of the region, in terms
of degree of rurality, cause of death, place of care (PoC), and PoD. The interview topics
were based around the patients and FCGs’ care needs, support, meaning of the ‘good
death’, and preferred PoC and PoD. The interview schedule and my approach deepened
over the course of the study, as questions were asked surrounding an issue raised by a
previous participant. Interviews were audio-recorded and transcribed verbatim. Data
were analysed using thematic analysis techniques.
1.3.4 Population death data survey
While it was essential to gain deep understanding of the ‘good death’ and PoD within
the Snowy Monaro region, it was also important that the data obtained through
qualitative methods were placed within a regional context, and on a background of
actual current practice. Unfortunately, I was denied access to the regional death data
held by the NSW Registry of Births, Deaths and Marriages (for reasons explained in
Chapter Four, Section 4.5.3) and therefore, death data were initially collected from the
local funeral director, local newspaper obituary and funeral notices, and funeral
announcements on local radio. As data collection began, it became apparent that data
relating to decedents residing in the Bombala area were missing, due to adjacent out-of-
region funeral directors being engaged, and lack of obituaries published in the once
weekly ‘Bombala Times’. An amendment was made to the ethics approval to recruit
additional informants, including the director of nursing (DON) or nursing managers of
12
local residential aged care (RAC) and inpatient health facilities. The websites of out-of-
region funeral directors were also scrutinised. Waiver of consent from decedents’ next
of kin was granted by ethics committees, on the basis that names and addresses were not
collected from the funeral directors and facilities, and the use of any data was de-
identified. Demographic data, including sex, age at death, marital status, last place of
residence, PoD, and cause of death, were collected over a 16-month period, for 224
deceased residents. It became apparent conflicting information was provided in some
obituary notices, with the last address of some long term RACF residents reported as
their previous private residence. Therefore, a further amendment was made to the ethics
protocol to permit DONs to provide the length of stay for their deceased residents,
identified through obituary and funeral notices. Data were initially charted descriptively.
Cross tabulations by PoD and key demographic data were performed. Bivariate logistic
regression was undertaken to identify factors that may influence the actual PoD.
1.4 Challenges specific to this study
There were several challenges, both ethical and practical, that required attention during
the preparation and data collection stages of this research study. Some issues were
addressed during the HREC application phase, and others, like the Bombala death data
mentioned in the previous section, required an amendment to the ethics approval. The
main issues (researcher bias and potential recruitment coercion), related to my dual
position of researcher, and palliative medicine specialist practicing within the Snowy
Monaro region. Specific strategies were implemented and are discussed in Chapter
Four, Sections 4.4. 1, and 4.7.
1.5 Thesis structure
This thesis is organised into four parts, comprising nine chapters.
Part 1: Research introduction
In Chapter One I have provided the background to this thesis, and my own world view.
As previously reported, one measure of the success of quality palliative care is to enable
people to die in their preferred place, with the literature suggesting this to be home for
most people. A death at home is considered by some advocates of palliative care to be
13
central to the concept of the good death. However, the current viewpoints regarding
PoD are mostly urban focused, and obtained from healthy populations or healthcare
professionals. In this chapter, I declare my personal and professional perspectives on
PoD, and my thought processes in developing the objectives of this research study.
Part 2: Research preparation, methodology and methods
Chapter Two sets out the background rural literature relating to the rural ‘good death’,
and rural EoLC and PoD. The chapter begins by describing the discipline of palliative
care, the concept of the good death, and the challenges in defining ‘rural’. This is
followed by a summary of the findings of my published scoping and systematic
literature reviews, and their relevance to rural residents’ perspectives on the ‘good
death’, PoD in rural palliative care, and rural EoLC. The methods utilised in conducting
these reviews are outlined in Chapter 4 (Section 4.2). These reviews demonstrate the
paucity of rural EoLC research, especially from the experiences and perspectives of
rural patients and FCGs.
Chapter Three describes the setting, both in terms of location and target population. As
each rural community is unique, it is important that a visual image is created for the
reader. This chapter describes the Snowy Monaro region: the location, demographics,
history, and current medical services. The target population is briefly introduced. The
chapter concludes with a discussion surrounding the challenges faced in recruiting
participants to end-of-life research.
Chapter Four introduces the research questions to be answered, and the methodology
and methods underpinning the research study. This chapter concludes with a discussion
surrounding the ethical issues relevant to the research: the role of the researcher,
researcher bias, transparency, imbalance of power, limitations of the study, and research
rigour.
Part 3: Results
Chapter Five describes the actual place of death of Snowy Monaro residents, between
February 1, 2015 and May 31, 2016. In total, 224 deaths were identified. The chapter
begins with a description of the socio-demographics, cause of death, and PoD for all
224 deaths. While the focus of this thesis is the palliative subset, a brief description of
14
the cause of death and PoD for the non-palliative subgroup (those who died
unexpectedly of an acute condition) is provided. This is followed by a description of the
data related to the ‘palliative subgroup’ (those residents who died of an advanced
chronic condition considered amenable to, but not necessarily receiving, palliative care).
Chapter Six describes the actual and preferred place of death for the 15 patients
identified through the interviews, and explores the meaning of, and issues surrounding,
place of care and death, as described by the participants (patients, FCGs, HCPs). The
chapter begins with a vignette, Gwen’s story, which highlights the issues faced by the
participants and the events that necessitated a move away from home. Like most of the
participants, home was Gwen’s initial preferred PoD; however, due to changing
circumstances, (disease progression, carer unavailability, significant symptoms, loss of
mobility, lack of 24-hours, seven-days-a-week professional support at home), home
became unsustainable, and her preferences changed. I then explore the actual and
preferred PoD for all 15 identified patients. The meaning of home is explored through
the experiences and perspectives of the participants. I explore the congruence between
actual and preferred PoD, and the congruence between preferences of patients, FCGs
and HCPs. While home was the preferred PoD for most, home was not the actual PoD
for the majority. Most accepted the change providing the new place was a ‘safe’ place.
This chapter argues that PoD is only one element contributing to a ‘good death’ and that
a ‘home death’ should not be a gold standard, or a measure of the success, of palliative
care.
Chapter Seven presents the findings relating to the participants’ perspectives on the
‘good death’. The chapter begins with Ivan’s story, which illustrates a ‘good’ rural
death at home on the farm. However, while dying at home added to the ‘goodness’ of
Ivan’s death, it wasn’t the only influence. This chapter explores the difficulty of
defining the ‘good death’ as each participant holds a different view. I then explore the
factors associated with a ‘good death’, as described by the participants. A ‘safe’ death
emerged as the central theme and essential component of a ‘good death.’
Chapter Eight explores the influence of rural residency on the ‘good death’, PoD, and
safety. The chapter begins with Marianne’s story, which highlights the factors that can
facilitate or hinder a ‘good death’ and dying in one’s preferred PoD. Each patients’
15
place of residence is described. This is followed by a discussion on how the rural
participants viewed themselves, compared to urban residents. I then explore the factors
associated with rural residency that impact negatively on the ‘good death’ and PoD
(travel distances, lack of local specialist services, local medical expertise, and allied
health support, lack of all-hours in-home formal support). Finally, the benefits of rural
residency are explored (personalised care, community support, friendships with HCPs,
lifestyle and convenience).
Part 4: Discussion and conclusions
Chapter Nine presents the discussion, conclusions and implications of the research
presented in this thesis. The findings of this original research are discussed within the
context of the existing literature. The limitations of this research study are discussed,
followed by my reflections on my role throughout the study. The implications for policy
and practice, and further research, are discussed.
16
17
PART 2: RESEARCH PREPARATION,
METHODOLOGY AND METHODS
18
19
Introduction to Part 2: Research preparation,
methodology and methods
Part 2 of this thesis contains three chapters. Chapter Two discusses the current rural
literature relating to place of death, preferred place of death, the concept of the good
death, and end-of-life care from the experiences and perspectives of rural residents. Two
systematic and one scoping literature reviews were undertaken, with the findings
presented in this chapter. These literature reviews were considered to be research in
their own right, therefore the review methods are reported in Chapter Four.
As each rural region is unique, Chapter Three is devoted to describing the setting for
this research study, the Snowy Monaro region of New South Wales, Australia.
This was a mixed methods research study, using systematic and scoping literature
reviews, ethnography, and a quantitative death data survey. Chapter Four states the
research objectives, and discusses each of the methodologies, and methods used to
collect, record and analyse the data.
20
21
CHAPTER TWO: LITERATURE REVIEW
Introduction
The primary objective of my thesis was to explore the concept of the good death from
the experiences and perspectives of rural patients, with a life-limiting illness, and their
family caregivers (FCGs), within the rural Snowy Monaro region of NSW, Australia.
The literature suggests the ‘good death’ is dependent on context and individual
perspectives. Therefore, the secondary objectives were to determine the influence of
place of death (PoD), and rural residency on the ‘good death’, and to explore the
meaning behind the PoD preferences and decision-making process.
This literature review was iterative, commencing in mid-2014, in the earliest stages of
my thesis, and continuing until the thesis was in the final draft. Initially, a broad view of
the literature was taken, and while many excellent studies relevant to the ‘good death’,
PoD and end-of-life care (EoLC) were identified, most were either urban focused or
failed to identify the rural perspective in mixed rural/urban geography studies. It
became apparent, that to find the rural perspectives, a more focused and systematic
approach to the literature was required. This culminated in the completion and
publication of one scoping review: Rural residents’ perspectives on the rural ‘good
death’: a scoping review (Rainsford et al., 2016); and two systematic literature reviews:
(1) Place of death in rural palliative care: a systematic review (Rainsford et al., 2016),
and (2) Rural end-of-life care from the experiences and perspectives of patients and
family caregivers: a systematic literature review (Rainsford et al., 2017). The scoping
and systematic literature reviews may be considered research in their own-right, yet
form a significant part of my overall literature review. Therefore, the methods utilised in
conducting these reviews are outlined in Chapter Four (Section 4.2), and the findings
are included in this chapter, rather than in Part 3, Results. To avoid duplicating the
detailed tables containing the extracted data, only a summary table of the studies
identified in each review, including the updates, is included in this chapter (Table 2.1,
2.2, 2.3, and 2.4). A copy of the published papers, including the full data extraction
tables, can be found in Appendix 6.
22
When I commenced this literature review, the most recent identified comprehensive
review of rural palliative care was published in 2009 by Robinson et al. An additional
rural palliative care systematic review was published by Bakitas et al. in October 2015.
Both reviews highlight that most rural research has been focused on rural palliative and
health care services (intervention studies, needs assessment, program planning and
evaluation, education and finances), and professional attitudes, education and practices.
While it is vital that research seeks novel ways to improve the accessibility to, and
quality of, palliative care to rural residents, gaps in the rural literature remain. There is a
paucity of studies documenting rural patients’ and FCGs’ experiences; the influence of
rural place and culture on patient experiences; understanding PoD in rural settings; and
the challenges of recruiting rural and remote participants to studies.
Chapter structure
This literature review begins by describing the discipline of palliative care (Section 2.1)
and the concept of the good death (Section 2.2), as these are fundamental to my research
topic. The lack of consensus regarding rural definition is acknowledged (Section 2.3) as
this contributes to the challenges faced in conducting and comparing rural research.
This is followed by a summary of the findings of my scoping (Section 2.4) and
systematic (Sections 2.5 and 2.6) literature reviews, and their relevance to my research
study. A brief comment is made regarding current Australian rural studies (Section 2.7).
This chapter concludes with a summary of the overall literature review findings (Section
2.8).
2.1 Palliative Care
The World Health Organisation defines Palliative Care as ‘an approach that improves
the quality of life of patients and their families facing the problems associated with life-
threatening illness’ (WHO, 2002, p. 84). As such, palliative care addresses physical,
psychosocial and spiritual needs, and mandates this approach is the right of every
person regardless of where they live:
Palliative care
• Provides relief from pain and other distressing symptoms.
• Affirms life and regards dying as a normal process.
23
• Intends neither to hasten nor postpone death.
• Integrates the psychological and spiritual aspects of patient care.
• Offers a support system to help patients live as actively as possible
until death.
• Offers a support system to help the family cope during the patient’s
illness and in their own bereavement.
• Uses a team approach to address the needs of patients and their
families, including bereavement counselling, if indicated.
• Enhances quality of life, and may also positively influence the course
of illness.
• Is applicable early in the course of illness, in conjunction with other
therapies that are intended to prolong life, such as chemotherapy or
radiation therapy, and includes those investigations needed to better
understand and manage distressing clinical complications (WHO,
2002, p. 84).
2.1.1 Origins of palliative care
During the Middle-Ages the term ‘hospice’, derived from the Latin hospitium
[hospitality] and hospes [guest, stranger or foreigner] (Doyle et al., 1999, p. vi), was
used to describe a place of rest for travellers and pilgrims to Europe and the
Mediterranean. These hospices were run by religious orders who offered hospitality to
travellers, and people who were ill or dying. During the 19th century, the hospice re-
emerged in the United Kingdom and France as institutions run by religious orders
caring for people who were terminally ill, as well as providing accommodation for the
incurable and destitute (Doyle et al., 1999; Palliative Care South Australia, 2016).
The modern day hospice dates from 1967 with the opening of St Christopher’s Hospice,
London, established by Dame Cicely Saunders (1918 - 2005). With advances in health
care and medicine, the focus of care was directed towards those who could be cured.
Saunders became deeply concerned about the suffering (physical symptoms including
pain, spiritual and psycho-social distress) of those who could not be cured. The term
‘palliative’ (from French palliatif, or medieval Latin palliativus, from the verb palliare
'to cloak'), used to describe the care provided to the terminally ill, was first used in 1975
by a Canadian Surgeon, Balfour Mount, to avoid the stigma attached to ‘hospice’ as a
24
place for the destitute (Palliative Care South Australia, 2016). Today the terms ‘hospice’
and ‘palliative’ are used at times synonymously, and at times differently in various parts
of the world. In Australia, ‘hospice’ usually denotes an inpatient facility and ‘palliative’
an approach to comfort-oriented care, aimed at maintaining quality of life.
The term ‘Palliative Medicine’ was adopted in Great Britain in 1987, when palliative
medicine was recognised as a sub-speciality of medicine to undertake ‘the study and
management of patients with active, progressive and advanced disease for whom the
prognosis is limited and focus of care is on quality of life’ (Doyle et al., 1999, p. 3).
Traditionally, palliative care has been provided by GPs, and while they are well skilled
(Mitchell, 2002), specialist palliative care multi-disciplinary services have been
established, mostly in highly urbanised areas, especially for complex cases. In
September 1999, the Australasian Chapter of Palliative Medicine was established within
the Royal Australasian College of Physicians. The Chapter has the responsibility for
providing training and continuing professional development for Palliative Medicine
Physicians and Specialists in Australia and New Zealand (Cairns, 2007).
2.1.2 Contemporary palliative care
Palliative care has traditionally been associated with care of those with end-stage
incurable cancer. Australia, consistent with all developed countries, is facing a huge
increase of an aging population (Productivity Commission, 2013) as the ‘baby boomers’
age, and more people are expected to die from a non-malignant disease. In 2015,
ischaemic heart disease was the leading underlying cause of death in Australia, followed
by dementia, cerebrovascular disease (including stroke), cancer of the trachea, bronchus
and lung, and chronic lower respiratory diseases (ABS, 2015). Internationally, the
challenge now for palliative care is to ensure all patients with any life limiting illness
(including organ failure, frailty, dementia and cancer), where life expectancy is 12
months or less, have access to quality EoLC (Mitchell et al., 2010; Bausewein, 2017).
Studies suggest that people living in rural and remote areas receive less than optimal
care because they have limited access to specialist multidisciplinary palliative care
services (Lynch, 2013) and therefore require special consideration. In Australia, there
are both national and state palliative care strategies addressing the current challenges,
with the aim of achieving 24-hour-seven-day access to specialist palliative care teams
25
either face-to-face, by telephone or more recently by telehealth (Chi & Demiris, 2015).
At a national level, the Australian Government funds a range of national palliative care
projects, primarily focusing on education, training, quality improvement and advance
care planning. National policies are population based (Palliative Care Australia, 2003)
with the goal to reduce the number of patients presenting to emergency departments and
dying in hospital. While this may be appropriate in urban settings, where there are the
options of inpatient hospice/palliative care and community based specialist teams, some
rural studies have suggested that rural community hospitals may be an appropriate place
for EoLC (Payne et al., 2004). When available in rural communities, residential aged
care facilities (RACFs) are significant providers of EoLC. With the doubling in the
Australian population aged over 65 years by 2050 (Australian Government Treasury,
2010), and the predicted increase in dementia and other chronic illnesses, RACFs will
become even more significant providers of EoLC (Australian Institute of Health and
Welfare [AIHW], 2015).
Each Australian state and territory is responsible for the delivery and management of
palliative care within its own jurisdiction. As the study is being conducted in New South
Wales (NSW), only this state’s plan will be described.
In 2012, the NSW Government committed an additional $35 million to improve access
to community care and support for people requiring EoLC, and their families and carers.
The NSW Ministry of Health (2012) developed a plan ‘to increase access to palliative
care (2012-2016)’ aimed at ‘honouring people’s choices about the end of their lives’ by
enhancing services, expanding the skills of multidisciplinary teams providing such care,
and involving families more directly in the provision of care. In introducing this plan,
the then Minister for Health and Minister for Medical Research, Jillian Skinner, stated
the outcome aim was for ‘more people to have a real choice about being able to die at
home in an environment of love and comfort provided by those closest to them’ (p. 4).
One of NSW Government funded initiatives is the ‘Last-days-of-life home support
service’ (NSW Ministry of Health, 2013). In 2013, the HammondCare Consortium,
comprising HammondCare, Sacred Heart Health, and Calvary Health Care Sydney,
successfully won the tender to provide palliative care home support packages
throughout seven Local Health Districts (LHDs) in NSW: Central Coast, Far West,
26
Murrumbidgee, Northern Sydney, South East Sydney, Southern NSW, and Western
NSW. Silver Chain provides services in other areas. These in-home packages can be
mobilised quickly and offered on a tailored basis.
The NSW Agency for Clinical Innovation (ACI), a clinician-led organisation, was
tasked with developing new models of care, aimed at improving patient outcomes, and
building workforce capacity. One of their agendas is to expanded community-based
palliative care services, especially in rural areas and for special needs populations, with
the aim to increase the rate of home deaths by expanding support for families and
carers. In December 2015, ‘The Blueprint for Improvement’ was launched (NSW ACI,
2015). This is an online resource to guide service providers and LHDs in constructing
their own localised model of care. At the time of completing this literature review, no
reports on the outcomes of the initiatives were available; however, it is predicted that
with increased support, providing it is tailored to the needs of individuals and
communities, rural residents will have genuine choice regarding their EoLC and PoD.
The goal is to increase the rate of home deaths, and therefore allow more rural residents
achieve a ‘good death’.
On the 12th June 2017, the NSW Premier, Treasurer, and Minister for Health announced
an additional $100 million for palliative care services, over the next four years, would
be funded in the 2017 - 2018 NSW State budget. This funding will go towards palliative
care training for 300 nurses and allied health staff, 300 scholarships for rural and
regional staff to enhance palliative care skills, an additional six palliative care
specialists in rural and regional areas, and two specialist positions to provide relief to
other specialists in rural and regional areas (NSW Government Media Release, 2017).
2.1.3 Advance care planning
With an aging population (with many living with chronic co-morbidities), advances in
healthcare technology capable of prolonging life without necessarily improving quality
of life, and increasing health care expenditure (NSW Government, 2012) further
Government strategies are being developed. Advance care planning (ACP) is one
process to help people formulate and communicate their healthcare preferences during
future incapacity (Palliative Care Australia, 2008). By documenting choices regarding
future treatment, preferred PoC and PoD, families and healthcare professionals can be
27
assisted in ensuring palliative care is relevant to individual patients. As such, the
Australian Government (Australian Health Ministries Advisory Council, 2011) has
developed a national framework to support programs that encourage ACP.
2.1.4 The quality of palliative care
In the preface to the 2004 WHO booklet ‘Palliative Care: the solid facts’, Dame Cecily
Saunders states:
Palliative care begins from the understanding that every patient has his or her own
story, relationships and culture, and is worthy of respect as a unique individual. This
respect includes giving the best available medical care and making the advances of
recent decades fully available, so that all have the best chance of using their time well
(p. 7).
While patients will often ‘fight to survive’ most will do so only if the extension of time
maintains dignity, and an acceptable quality of life. However, quality is subjective and
based on an individual’s experience, expectations and values. In 1999, Singer et al.
suggested five domains to measure the quality of dying (adequate pain and symptom
management, avoiding inappropriate prolongation of dying, achieving sense of control,
relieving burden, strengthening relationships with loved ones). Two years later, Patrick
et al. (2001) argued quality death and dying is the congruence between a person’s
preferences for dying and the moment of death, and the reality of circumstances, as
reported by others. They suggest quality is based on adequate pain and symptom
control, and patient autonomy, and developed a 31-item model measuring the quality of
death and dying (QODD). This model, the most widely published and validated multi-
item measure available (Hales et al., 2010, p. 127), is based on six conceptual domains:
symptoms and personal care, preparation for death, moment of death, family, treatment
preferences, and whole person concerns (Patrick et al., 2001, p. 717).
While maintaining patient focus, the Australian and New Zealand Society of Palliative
Medicine (ANZSPM), an organisation for palliative medicine physicians, specialists,
trainees, general practitioners and other medical practitioners with an interest in
palliative care, have developed a document for Quality EoLC assessment and planning.
They consider five domains of EoLC essential:
28
1. Symptom management (holistic & multidisciplinary)
2. Advance care planning
3. Carer support
4. Coordination and integration of care
5. Terminal phase (last 7-10 days of life) - timely recognition of the ‘dying
phase’ and developing documented ‘Terminal Phase Plans’ to adequately
manage patients in their terminal phase and support their families, loved ones &
carers (ANZSPM, 2014, pp. 7-8).
The literature suggests one endpoint of quality palliative care is a ‘good death’;
however, Proulx and Jacelon (2004) suggest that a ‘good death’ cannot be measured
quantitatively. A ‘good death’ implies there can also be a ‘bad death’; however, who
decides what is ‘good’, as one person’s death is viewed through so many different
lenses - the patient, the family/carer, the doctor, the community/palliative care nurse, the
community and everyone else involved (Holdsworth, 2015; Meier et al., 2016).
Regardless, the ‘good death’ is a central palliative/hospice concept and will be explored.
2.2 The ‘good death’
2.2.1 Evolution of the concept of the good death
While death is inevitable as we are all “fellow- passengers to the grave” (Dickens, A
Christmas Carol; Stave 1: Marley's Ghost), achievement of a ‘good death’ is one of the
objectives, if not the sole aim, of palliative care. Kellehear (2008) argues that the term
‘good death’ has ‘two commonly cited derivatives’. The first is derived from the Greek
words eu thanatos from where the word ‘euthanasia’ is derived; eu meaning "good" and
thanatos meaning "death", that is to die well, or to die suddenly and gently, painlessly,
and easily. The second Greek derivative, more aligned with the palliative care
movement, is kalos thanatos, that is to die ‘beautifully or in an ideal or exemplary
way…not sudden …but well prepared by the dying person…A dying that conforms to
the wider community expectation of making death as positive and meaningful as
possible to as many people as possible’ (Kellehear, 2008, p. 90).
Philippe Aries, often considered a pioneer in the writing of social history, published a
set of lectures in which he reviewed ‘Western attitudes toward death from the middle
29
ages to the present [early 1970s]’ (Aries, 1975). Kellehear (2008) and Aries (1975)
describe the changes over time in the concept of the good death, and their arguments are
summarised in the remainder of this sub-section.
In pre-modern times, while dying was expected at a younger age compared today, the
manner and timing of death was often sudden and unpredictable, usually the result of
accident, war, infection, or childbirth. Personal awareness of death was brief, as the
dying person, for example, bled to death. Kellehear (2008) argues that much of the
Stone-Age dying process was considered to take place ‘after death’, as the deceased
passed from one world to the next, and transitioned from one social role to another, for
example father to ancestor, or chief to god, as depicted in cave drawings. This after-life
otherworld journey, thought to be at the time treacherous, was undertaken alone and
without support of family and friends. Inheritance was the grave goods supplied by the
survivors, to assist in a safe journey, or to prevent the deceased from returning and
tormenting the survivors. Farewells were possibly ambivalent as it was uncertain if the
dead would return as a protector or as a ‘haunting ghost’. It is difficult to know what
was considered a ‘good death’ in these times; however, it seems the dying person
played a minor role, if any role, with Kellehear (2008) suggesting a self-awareness of
imminent death, albeit very brief, or the quality of the grave gifts provided by the family
and community were the influencing factors.
The concept of the good death is dynamic, having evolved over time in pace with
urbanisation and health care technology (Aries, 1975; Kellehear, 2008). Over time, the
concept of the good death progressed from having a self-awareness of impending death,
to having the ability to self-prepare for death (the ‘tamed death’ Aries 1975); followed
by a ‘managed death’ (Howarth, 2007), whereby the doctor, priest (or equivalent) and
lawyer became important in assisting the dying person to ensure all final tasks
(physical, emotional, financial, spiritual and social) were completed. Today, this could
include the preparation of an advance care plan and/or directive, or the appointment of a
surrogate decision maker (enduring power of attorney) who can make health care or
financial decisions on the dying person’s behalf if they are unable to voice their own
preferences. This allows the dying person to manage their own death in advance.
Once only an ‘otherworld’ experience, death became a ‘this world’ experience with a
shift from the deceased being the beneficiary of grave gifts to the one who bequeathed
30
inheritance to the survivors. The otherworld journey became more of a moral issue as
death became ‘a ritual organized by the dying person himself, who presided over it and
knew its protocol’ (Aries, 1975, p. 11). Depending on religious beliefs, the pre-modern
Western death required the dying person to have time to brood over the sadness of
dying, pardon those close to him, confess his sins to God and ask for absolution. Having
lived a good life, provided a suitable inheritance for survivors, taken opportunity to put
things right and ensuring one’s world was not left in chaos (Kellehear, 2008, p. 88)
helped secure a safe passage for the deceased.
In comparison, according to Kellehear (2008, pp. 92-93), a ‘bad’ death involved a
sudden or violent death (including suicide), dying alone or away from homeland, dying
young or in childhood, dying without an heir or not having a proper burial. Lack of self-
awareness of one’s impending death and the inability to prepare a smooth transition of
material wealth to one’s heirs was also considered a ‘bad’ death.
Dying was once a public affair, occurring at home and supported by the community.
Over time, dying has become increasingly private and medicalised, hidden away in
institutions, the ‘forbidden’ death (Aries, 1975) or ‘death that is seen not to happen’
(Walters, 2004, p. 405), such as in hospitals or nursing homes (Gomes & Higginson,
2008) despite studies indicating home now has again become the preferred place
(Gomes et al., 2013).
2.2.2 Contemporary ‘good death’
The contemporary ‘good death’, at least in the developed world, is what Walters (2004)
describes as a ‘controlled death’. With medicalisation of dying, advances in medical
technologies and the ability to prolong life, without necessarily maintaining quality of
life, the ‘good death’ is now one of the main aims of the palliative care/hospice
movement. However, some argue the contemporary notion of the ‘good death’ is
idealised, ‘limits spontaneity’ (Cottrell & Duggleby, 2016) and puts ‘pressure on a
dying person to be a ‘good patient’ (Proulx & Jacelon, 2004, p. 116). While death is a
normal end to life (WHO, 2015), one could argue that many deaths are not entirely
‘good’ due to the nature of the illness (Holdsworth, 2015), the age of the dying person
(Counts & Counts, 2004; van der Geest, 2004; Huy, 2007; Dilger, 2008) and individual
persons’ perspectives. The goal should be to achieve the ‘best possible death given the
31
circumstances’ (Holdsworth, 2015, p. 839). McNamara (2004) argues the term ‘good
death’ be replaced with ‘good enough’ death, due to the frequent disparity between
patient, family, and health providers’ points of view. As palliative care becomes
‘individualised and eclectic’ (ibid, p. 929), consensus regarding the ‘goodness’ of a
death is not always possible. Today, most people expect to ‘die very much in the same
way they have lived’ (ibid, p. 930), thus satisfying realistic expectations of the dying
person (Raisio et al., 2015).
The move towards autonomy and control, particularly in Western medicine, has created
tension between stakeholders who often hold different priorities and expectations
towards end-of-life care (Holdsworth, 2015). A study by Steinhauser et al. (2000)
reported consistent high ranking of pain and symptom management, preparation for
death, achieving a sense of completion, decisions about treatment preferences, and
being treated as a ‘whole person’ across all stakeholders (patients, families, and care
providers including clergy and volunteers). However, this same study reported
considerable variation between stakeholders in regards to place of death, being mentally
aware, not being a burden, having funeral arrangements in place, and coming to peace
with God. While EoLC must reflect the needs of the dying person and their carers,
ideally it must also reflect the needs of the community (Cottrell & Duggleby, 2016). In
1996, Kearl argued that death becomes ‘good’ when it also serve the needs of survivors
– the family and the wider community. While there are strong similarities between
countries in regards the ‘good death’, in developing countries ‘poverty shaped how
people died’ (Grant et al., 2011, p. 5) and beliefs and rituals surrounding dying remain a
mix of tradition and contemporary.
2.3 Challenges in defining rural
As the focus of my research study is to explore the concept of the good death within one
rural context, it is important to find the voice of rural patients and FCGs (Bakitas et al.,
2015). To ensure palliative care services are ‘embraced by community members’ (ibid,
p. 462), and ‘to remain socially relevant, end-of-life care ideally must reflect the needs
of the dying individual…. within diverse cultural and geographic areas’ (Cottrell &
Duggleby, 2016, p. 26). Studies consistently report that rural folk perceive themselves
as different to urban residents and clearly hold distinct views on what it means to die
32
well (Wilson et al., 2009a, 2009b; C Robinson et al., 2010; Spice et al., 2012). Gessert
et al. (2015) reported independence, stoicism, and fatalism to be common traits found in
rural residents. Compared to urban residents, there is sometimes greater pressure placed
on residents living in small communities, to conform to social expectations and norms.
However, rural researchers are challenged in their search for commonalities. Firstly,
despite an increase in rural palliative care research, including in Australia (C Robinson
et al., 2009), studies remain limited. Secondly, and possibly more importantly, there is
no universal rurality index or international consensus on rural definition. With a lack of
homogeneity of health services and cultural context, it is difficult to compare results,
not only between countries, but also within countries. This challenge is greater when
comparing studies from developed and developing countries where priorities vary
significantly.
The definition of rurality ‘is a multifaceted concept … and frequently relies on
stereotypes and personal experiences’ (Hart et al., 2005, p. 1149). Definitions have been
based on population size, population density or demographics; distance from urban
centres and services; or defined as a specific ‘culture’ (ibid 2005). For example, since
the early 1990s four different geographic classifications have been developed in
Australia (AIHW, 2004): Rural, Remote and Metropolitan Areas (RRMA)
classification, Accessibility/Remoteness Index of Australia (ARIA), Australian
Standard Geographical Classification (ASGC), and Modified Monash Model (MMM).
The rural town of Cooma, (population 6742 [August 2016], 115 kms from the nearest
major urban area) provides an example of conflicting rural definitions. Depending on
which classification is applied, Cooma is considered ‘rural’ based on population size
(RRMA) or ‘inner regional’ based on the road distance to accessible services (ASGC).
The classification has significant implications on funding, service provision and
financial incentives to attract general practitioners to areas of need. In the rural studies
identified in my PoD systematic review, rural populations ranged from <1000 (Burge,
2005; Papke & Koch, 2007); to <9999 (Goodridge et al., 2010b); with rural towns
ranging from 2000 (Escobar Pinzon, 2011) to <28,000 (Herd, 1990). Rural areas
typically have fewer health and social services as compared to urban areas (Downing &
Jack, 2012) with some included studies described as having access to large county
33
hospitals (Axelsson & Christensen, 1996), inpatient hospices and residential aged care,
while others had limited medical facilities (Yasumura et al., 2000).
Notwithstanding the limitations in the current rural literature, rural people believe they
have some unique perspectives on and concerns about dying, end-of-life care and the
‘good death’. While it could be argued that the concerns raised by rural participants are
not dissimilar to urban dwellers (desire for autonomy and good quality EoLC), a
narrative review of eight studies by Kirby et al. (2016, p. 1) described rural residents as
‘more accepting of death and less likely to intervene to delay death’ and concluded that
the ‘palliative needs of rural and remote residents are related to context ….and are
shaped by reduced access and availability of services’. This is supported by the findings
of my systematic literature review, which identified substantial differences in the
provision of care due to distances and the reality that sometimes people must move
away from their community to receive the care required.
Rural people have a deep concern for their community and its members (Kirby et al.,
2016). Strong relationships and informal community services are important factors that
facilitate care within rural communities. Pesut et al. (2010, p. 190) reported in their
study that:
Participants spoke eloquently of the benefits of their rural lifestyle including physical
beauty, privacy, and accessibility of recreational activities. The level of support
provided by community members was an important factor in why individuals valued
rural life.
It is this value of rural life that contributes to the quality of EoLC, as death and dying is
an individualised experience with cultural, religious and political values and beliefs
influencing the quality of dying (Cottrell & Duggleby, 2016). Facilitators and barriers to
receiving EoLC in rural/remote settings, and the influence of rural place and culture on
EoLC experiences were two objectives of my scoping review.
34
2.4 Rural residents’ perspectives on the rural ‘good death’: a scoping
review
As discussed previously, many of the viewpoints and research findings on the ‘good
death’ reflect the urban voice (Robinson et al., 2009; Bakitas et al., 2015); however,
rural folk perceive themselves as different to urban residents (Wilson et al., 2009a,
2009b). A scoping literature review was undertaken exploring rural residents’
perspectives. The methods are described in Chapter Four (Section 4.2.1). The literature
search, conducted from 2nd January 2016 to 28th February 2017, identified 20 papers,
reporting on 17 studies and one systematic review (Table 2.1). A copy of the published
paper, including tables containing the data extracted from the initial identified studies
can be found in Appendix 6. A summary of the findings is included in this section of
my literature review.
The included articles were published between 2000 and 2015, with data collected
between 1981 and 2013. Rural perspectives were reported in articles from both
developed and developing countries. While all studies were described as rural, there was
no standardised definition of rurality. Locations were heterogeneous. All the eligible
studies were qualitative in nature, using focus groups, interviews, written surveys,
clinical encounters and community observations. The exact number of participants is
unknown as two anthropological studies included whole communities (Counts &
Counts, 2004; Easom et al., 2006). Of the 751 identified rural participants, 84 were rural
patients, 68 rural FCGs, 323 rural HCPs, 153 rural residents, 83 rural community
leaders and 40 rural health administrators and policy makers. Of the identified
informants 20% were patients or FCGs and of these 76% were African. HIV/AIDS
(considered a stigmatised disease) or cancer accounted for most terminal illnesses in
Africa. Of the 19 patients in developed nations, 17 had cancer, one dementia and one
cerebral vascular disease.
The concept of the good death was widely acknowledged in both developed and
developing countries, and referred to the death event, the dying process, the meaning of
death or the after-death concept. The identified studies illustrated that the ‘good death’
is subjective, with priorities based on personal, cultural, social and religious
perspectives. However, despite the diversity, common themes were identified. The
35
themes surrounding the contemporary rural ‘good death’ included being pain free,
maintaining dignity and autonomy in decision making, having support of family, and
providing opportunity for the dying person to “sort out” personal affairs (Raisio et al.,
2015; Davies et al., 2016; Meier et al., 2016).
Table 2.1: Studies included in rural 'good death' literature review
First
Author
Publication
Date
Country Method Diagnosis Informants Quality
Arnaert 2009 Canada Interviews
(ITV), focus
groups (FGD)
- 5 Homecare
nurses
Medium
Beckstrand 2012, 2015 USA Written
survey
- 236 Nurses Medium
Biggs 2014 Canada Semi-
structured
ITV
Dying 10 Dying
Patients,
8 FCGs
High
Counts 2004 Papua &
New Guinea
Field work;
Observations
- Community Low
Cruickshank 2010 Scotland Patient- ITV;
FCG-
Bereavement
nurse- FGD
- 4 Patients,
8 FCGs,
12 nurses
Medium
Devik 2013 Norway Narrative ITV Cancer 5 Cancer
patients
Medium
Dilger 2008 Tanzania Field work HIV/AIDS HIV patients Low
Easom 2006 USA Mixed:
Written
survey; ITV
- 9 Nurses Low
Felt 2000 USA FGD - Community Medium
Grant 2003 Kenya Semi-
structured
ITV
Cancer,
HIV
32 Patients
(cancer,
AIDS),
24 carers
High
Grant 2011 Uganda,
Kenya,
Malawi
ITV - 33 Patients,
27 FCGs,
36 HCP, 25
volunteers,
29 leaders
(community)
Medium
Gysels 2011 Africa Systematic
review
- -
Huy 2007 Vietnam FGD - Farmers Low
Joarder 2014 Bangladesh ITV Elderly 8 Elderly
villagers
Medium
Knight 2014 England ITV - 4 Clergy Medium
Van der
Geest
2004 Ghana Fieldwork
observation,
ITV, FGD.
- Farmers Medium
Veillette 2010 Canada ITV, FGD - Community High
Wilson 2009a,
2009b
Canada ITV, FGD. - Community,
HCP, policy
makers
High
Abbreviations: ITV: interviews; FGD: focus group discussions; FCG: family caregiver;
HCP: healthcare professionals; (-) details not provided; informant number reported when provided.
36
Consistent with the WHO mandate (2002, p. 84), and urban studies (Holdsworth, 2015;
Raisio et al., 2015; Davies et al., 2016; Meier et al,. 2016), rural residents identified five
dimensions considered important for facilitating a ‘good death’– physical, emotional,
social, spiritual, and cultural. The dominant theme, from both developed and developing
countries was that a rural ‘good death’ is one that is peaceful, free of pain and without
suffering. The themes describing the rural ‘good death’ parallel the urban view, and
include a ‘controlled’ death, with control over symptoms, place of death, decision
making, manner of death, and independence (Felt et al., 2000; Counts &Counts, 2004;
Dilger, 2008; Wilson et al., 2009a, 2009b; Veillette et al., 2010; Knight, 2014); a
‘timely’ death (Counts & Counts, 2004), that is a death coming ‘naturally and after a
long and well- spent life’ (van der Geest, 2004, p. 899) and ‘hopefully in my sleep’ (Felt
et al., 2000, p.405) after having had opportunity to say goodbye to family; a ‘dignified’
death by maintaining identity, self-worth, integrity and control (Wilson et al., 2009a;
Devik et al., 2013); a ‘social’ death, such as to die within the community with family
present (Wilson et al., 2009a, 2009b); and a ‘noble’ death such as through enduring the
situation (Grant et al., 2003; Devik et al., 2013). Two articles (van der Geest, 2004;
Knight, 2014) acknowledged the difficulty of defining a ‘good death’, as it is dependent
on individual interpretations, perspectives and priorities.
Place of death was considered in terms of geographical location (community), and
physical space (home, hospital, residential aged care). The strong connection to one’s
rural/remote community meant dying within one’s rural locality was identified as a
critical element in achieving a ‘good death’. While the literature reports that PoD, in
terms of physical space, is one measure of a contemporary ‘good death’, a number of
authors suggest that place is only one factor, and a ‘home’ death is not universally or
necessarily the most important determinant of a ‘good death’ (Biggs, 2015; Hoare et al.,
2015; Davies et al., 2016). If it is not possible to die at home surrounded by family, then
it is important to die within the rural community (Wilson et al., 2009a, 2009b; Veillette
et al., 2010; Biggs, 2015) as ‘the good rural death is an outcome of rural community
values’ where ‘some would prefer to have less care and fewer services’ (Veillette et al.,
2010, p. 163) than leave their community.
37
While 20 papers were reviewed, due to heterogeneity across locations, cultures,
participants and research studies’ objectives, there is currently insufficient data to
generalise rural residents’ perspectives and what it means for them to ‘die well’.
2.5. Place of death in rural palliative care: a systematic review.
In 2004, the World Health Organisation (WHO, 2004, p. 17) released a report stating
that meeting individual preferences for place of care and death should be the ultimate
measure of the success of the palliative care provided. In doing so, the WHO urged
policy makers to encourage health services to inquire of people their preference for, and
monitor the wishes of seriously ill people concerning place of care and death, and
routinely monitor PoD as one interim measure of success. The inference being that
congruence between preferred and actual PoD is a fundamental component of the ‘good
death’.
Cross-national studies in developed countries (Cohen et al., 2010) report most people,
including Australians, (Howat et al., 2007; Swerrisen & Duckett, 2014) die in hospital.
In 2013, Gomes et al. published a systematic review reporting that home was the
preferred place of care and death for 31-87% of patients, 25-64% of carers and 49-70%
of the general public; median percentage 54% (Broad et al., 2013). Some studies report
a shift in deaths away from hospital (Wilson et al., 2009c) towards home (Burge et al.,
2003) and care homes (Houttekier et al., 2011), while others report the opposite
(Mystakidou et al., 2009; Tang et al., 2009; Yamagishi et al., 2012) with a prediction
that by 2030 home deaths in the UK will have decreased by 42% and institutionalised
deaths increased by 20% (Gomes & Higginson, 2008).
As my research study is rural focused, it is important to determine if preferred and
actual place of care/death are different for rural residents. Healthcare resources,
including palliative care, are less available in rural areas, with reports indicating rural
residents are less likely to die at home than urban residents. This could imply that rural
residents are also less likely to have a ‘good death.’ However, rural residents’
preferences for PoD are not clearly reported. A systematic review was undertaken to
review the rural data. The methods used for this review are described in Chapter Four
(Section 4.2.2.1). A copy of the published paper, including the original data extraction
38
tables, can be found in Appendix 6. A literature search, conducted from 1st September
2014 to 28th February 2017, identified 44 papers, reporting on 39 separate studies. All
studies were included in this review, irrespective of their quality (Table 2.2, Table 2.3).
The rural studies were heterogeneous in terms of rural definition and participants, with
wide variations reported in rural home death rates. Overall, home death rates ranged
from 12% in Sweden (Axelsson & Christensen, 1996) to 81.7% in Taiwan (Lin & Lin,
2007), with cancer patients having the greatest chance of dying at home (Costantini et
al., 2000; Lin & Lin, 2007). Three Australian rural studies reported home death rates of
18.1% (Burns et al., 2015), 19% (Howat et al., 2007), and 26.7% (Crawford, 2000).
Except for two UK studies (McCall & Rice, 2005; Thomas et al., 2003, 2004), more
than 50% of rural participants expressed a preference for dying at home. Home was the
preferred place for 25% of participants (cancer patients) in the two UK studies;
however, Thomas, Morris and Gatrell (2003) reported an additional 33% of participants
selected ‘home or hospice’ as their preferred place, an option not available in other
studies. Consistent with urban dwellers, rural residents wanted to die at home,
surrounded by family and friends (Veillette et al., 2010) and in the case of Indigenous
Australians, connected to their land and family (McGrath, 2007); however, most deaths
occurred away from home. While hospital was the commonest PoD in some studies,
(New Zealand 21.5% (Smyth et al., 2010); Canada 76.2% (Burge, 2005; Burge et al.,
2003), a study in rural South West Scotland (Black et al., 2016) reported PoD was
significantly dependent on cause of death. In Black’s study, between 2000 and 2010
there was a decline in home deaths and an increase in the number of deaths occurring in
institutions; however, most cancer deaths occurred in the specialist palliative inpatient
unit within acute hospitals. Increasing age was a strong negative predictor for a home
death, with RACFs the commonest place of death for patients with dementia.
39
Table 2.2: Studies included in actual place of death literature review
First Author Publication
Date
Country Informant Diagnosis PoD Home death Quality
Aabom 2005 Denmark Death certificates Cancer Home vs institution 29% Medium
Axelsson 1996 Sweden Death certificates Cancer Home, hospital, RACF 12% High
Black 2016 Scotland National mortality register. Palliative Acute hospital, cottage hospital,
RACF, home.
All deaths: 25%
Cancer 27.8%
Dementia 9.5%
Respiratory 17.6%
IHD 43.7%
Medium
Burge 2003; 2005 Canada Death certificates Cancer Hospital vs out-of-hospital 23.8% High
Burns 2015 Australia Bereaved relative ITV Terminal illness Hospital, home, hospice, RACF 18.1% High
Carroll 1998 Scotland Single practice medical
records
Cancer Home, community hospital,
residential home, hospital or
hospice
30% Low
Castillo-Guzman 2013 Mexico National records Cancer Home, hospital, other 69.4% Medium
Catalan-Fernandez 1991 Spain Relative face-to-face Q Cancer Home or hospital 65.9% Medium
Cohen/
Houttekier
2008/
2010
Belgium,
Netherlands,
England
Death certificates Cancer Home vs outside home Belgium 28.2%,
Netherlands 39.1%,
England 20.7%
High
Constantini 2000 Italy Regional mortality register Cancer Home or hospital (incl. RACF) 53.4% High
Crawford 2000 Australia Death certificates Cancer Home or hospital 26.7% Medium
Escobar Pinzon 2011a
2011b
Germany Relative; written survey All Home, hospital, palliative care
facility, aged care home,
elsewhere
43% Medium
Gatrell 2003 England Mortality data Cancer Home, hospital, hospice,
elsewhere, NH/retirement home
aOR 1.334
95% CI 1.139-1.562
Urban home=1
High
Goodridge 2010b Canada Service files COPD or
lung cancer
Home, hospital, aged care 15.4% Medium
Herd 1990 England FCG ITV Cancer Home, hospital 53% High
40
Table 2.2 continued
First Author Publication
Date
Country Informant Diagnosis PoD Home death Quality
Higginson 2017 England Death records COPD Hospital, home, hospice
(inpatient), nursing home, care
home/residential home,
elsewhere.
Rural hospital COPD:
OR 0.93; 95% CI 0.92-
0.94
Urban hospital =1
Medium
Howat 2007 Australia PC service records Palliative care Home, hospital, RACF 19% Medium
Hunt 2001 Australia Cancer registry Cancer Hospital, hospice, Home, NH aOR 3.5 (95% CI 3.03-
4.04)
urban home=1
High
Jayaraman 2013 Canada Death certificates All Home, extended care facility,
hospital or other.
20.2% High
Lavergne 2015 Canada Linked health and census
data
Palliative In, or out, of hospital. RACF
excluded
26.6%
out-of-hospital
Medium
Lazenby 2010; 2012 Botswana Death certificates All Home or hospital 38.3% Medium
Lin 2007 Taiwan Death certificates cancer Home or hospital U6:78.4%
U7:81.7%
High
Lopez-Campos
2013 Spain Death certificates COPD Home, hospital, hospice, other
aOR 2.8 (95% CI 2.46-
3.16)
urban home =1
Medium
McNamara 2007 Australia Death certificates Palliative UPoR, hospital, RACF,
hospice. Other
Rural aOR=0.74 (95%
CI 0.61-0.91); Remote
aOR=1.03 (95% CI 0.75-
1.41)
City UPoR=1
Medium
Papke 2007 Germany Death certificates Cancer Home, hospital, RACF Small town: aOR 1.54
(95% CI 1.29-1.82)
Large town home=1
Medium
Smyth 2010 NZ GP postal survey Palliative care Home, hospital, hospice,
RACF
46.2% Low
Thomas 2003 England FCG/HCP bereavement
ITV
Cancer home, hospital, hospice,
elsewhere
25% (n=8) Medium
Yasumura 2000 Japan FCG ITV Palliative Home, hospital 47.7% High
41
Table 2.3: Studies included in preferred place of death literature review
First Author Date Country Method Diagnosis Informant Study Home PPoD Quality
Carroll 1998 Scotland ITV/ Medical
notes audit
Cancer GP Retrospective PPoD:77% (audit) Medium
Champion 2015 Australia Case note review
Palliative Notes Retrospective 100% Low
Choi 2005 South Korea Survey Cancer Patients/ FCG Prospective Patient 47%;
FCG 51%
Medium
Foreman 2006 Australia State Survey - Population Prospective 71.1%
Medium
Fukui 2011 Japan National survey - Population Prospective 51% (PoC)
Medium
Gu 2014 China FTF survey Cancer Patients/ FCG Prospective 60.4%
Medium
Howell 2011 Canada FTF ITV Advanced
disease
Patients Prospective 80% High
McCall 2005 Scotland FTF ITV Cancer Patients Prospective 25% (PoC)
High
McGrath 2007 Australia FTF ITV Palliative Patients, FCG,
HCP
Prospective - Medium
Shih 2015 Taiwan Postal Q - Population Prospective 79%
Medium
Thomas 2003, 2004 England FTF ITV Cancer Patients Prospective 25%; home or
hospice 33%
High
Veillette 2010 Canada ITV, FG - Community Prospective,
retrospective
- High
Wilson 2013 Canada Telephone survey - General public Prospective 74.9%
Medium
Abbreviations for Table 2.2 and Table 2.3: PPoD: preferred place of death; GP: general practitioner; FCG: family caregiver; HCP: healthcare professional;
PoC: place of care; ITV: interview; FTF: face-to-face; Q: questionnaire; FG: focus group; (-): item not reported.
RACF: residential aged care facility; IHD: ischaemic heart disease; COPD: chronic obstructive pulmonary disease; UPoR: usual place of residence; OR: odds ratio
42
Compared to urban residents, there were inconsistent experiences between rural
locations. Rural residents in South Australia had a greater chance of dying at home
[aOR 3.5; 95% CI 3.03, 4.04] (Hunt et al., 2001) than urban residents. Likewise, in a
study by Higginson et al. (2017), rural patients with end stage COPD were less likely to
die in hospital than patients residing in urban locations (OR0.93, 95% CI 0.92, 0.94).
However, a Canadian study by Lavergne et al. (2015) reported no significant difference
in the odds of dying in hospital for urban and rural residents.
The qualitative studies provided meaning to the actual PoD. Deciding on PoD was
found to be a complex process, in which wishes and preferences were not necessarily
the same. While wishing to remain home, the participants in the study by McCall and
Rice (2005) accepted a home death may not be possible, due to ‘concern about their
carer and not wanting to become a burden’ (p. 543). For these patients, hospital became
their preference, and possibly a substitute for inpatient hospice care.
Factors influencing PoC/PoD in rural settings were not clearly identified in the included
studies, but when reported, include patients’ functional status and clinical condition,
carer and social networks, and health-system facilities. The uniqueness of individual
rural communities was evident. Locations with stronger traditional cultures and values
exhibited the highest numbers of home deaths, or the strongest preferences for home
deaths; however, these locations often had less access to hospitals, hospices and
RACFs.
As previously stated, the congruence between preferred and actual PoC and PoD has
been mooted as a quality outcome for palliative care (Patrick et al., 2001), with health
policies aimed at enabling people to die in their preferred place (Dept. Health [UK],
2008; NSW Ministry of Health, 2012; Hoare, 2015). Based on the current literature, the
assumption is that for most people, the preferred PoC and PoD is home, being in
‘familiar surroundings in the company of close family and/or friends’ (Dept. Health
[UK], 2008, p. 9). With only two studies reporting the actual and preferred place of
death, there is insufficient data to draw any conclusions as to the degree of congruence
within rural settings.
43
One limitation of most studies included in the systematic review, was that they were
either population surveys reporting hypothetical preferences of healthy people, or
retrospective bereavement interviews/questionnaires of families and carers. Few studies
collected preferences directly from patients, explored the impact of incongruence in
preferred and actual PoC/PoD on patients and FCGs, or addressed the reliability of the
retrospective informants. A non-rural study by Agar et al. (2008), suggests PoC is not a
‘euphemism’ for PoD. McCall and Rice (2005) reported that their rural patients changed
their preferences as illness progressed. A recent non-rural study by Hoare et al. (2015)
suggests that it is not possible to state accurately what proportion of cancer patients
prefer to die at home due to the extent of missing data, as studies have frequently
omitted to report preferences from those people with no preference or who have not
been asked. Other non-rural studies report PoD is not necessarily a dominant concern of
the dying person (Steinhauser et al., 2000) and a hospital death is not necessarily a ‘bad’
death (Gott et al., 2004). Meanwhile, Menec et al. (2010) suggest rural residents are
disadvantaged by the lack of inpatient facilities, and that rural hospitals act as a
substitute for inpatient hospice.
Due to the limitations of the current rural studies, it remains unclear if the ‘push’ to
increase the rate of home deaths in rural settings is justified, or even if it is appropriate
to use achievement of preferred PoD as a measure of the quality of palliative care, and
the ‘good death’. While Seal et al. (2014) reported some FCGs experienced adverse
effects when their family member was unable to die at home, in their preferred place, a
recent Japanese family bereavement study (Miyajim et al., 2014) reported that failing to
die in a patient’s favoured place was not a contributing factor in complicated family
grief. This Japanese study suggested that having come to terms with life’s unanswerable
questions, being surrounded by family and community, and quality bereavement
support were more appropriate outcome measures. Providing a person’s universal need
for security, significance and self-worth are met at the end of life, then being in
relationship and surrounded by significant people may be more important than the PoD.
Robinson et al. (2010) suggest that a death at home is not necessarily the most
appropriate or desirable place in rural settings, due to often limited palliative care
resources, especially after-hours. Current studies are reporting on the significant, and
often overlooked, burden and cost (financial, physical, and emotional) to FCGs caring
for a family member at home (Gott et al., 2015). Despite informal support of rural
44
communities, travel distances and limited resources often increase the burden and cost
to rural FCGs. Patients may only be able to die well, and have a ‘good death’ if they
leave their home or farm to die in the rural hospital.
While achieving one’s preferred PoC and PoD may appear to be an easy outcome to
measure, an easy box to tick on data collection, it would be helpful to explore the
meaning behind, and the significance placed on the decisions made by patients, and the
experiences and perspectives of rural patients and FGGs in end-of-life care.
2.6 Rural end-of-life care from the experiences and perspectives of
patients and family caregivers: a systematic literature review
The second systematic literature review aimed to describe the EoLC experiences of
rural patients and FCGs; to identify facilitators and barriers to achieving a ‘good death’
in rural/remote settings; and to describe the influence of rural place and culture on
EoLC experiences. The methods for this systematic review are described in Chapter 4
(Section 4.2.2.2). The literature search, conducted from 2nd January 2016 to 28th
February 2017, identified 32 articles, reporting on 26 separate studies from both
developed and developing countries, and one systematic review (Kirby et al., 2016). A
copy of the published paper, including the original data extraction tables, can be found
in Appendix 6. A summary of the findings, including a summary table of the studies
identified in the original and updated literature searches (Table 2.4), is provided in this
section of the chapter.
Consistent with my two previous literature reviews, heterogeneity between studies was
demonstrated in rural definitions, settings and degree of isolation. Three studies (Lockie
et al., 2010; Johnston et al., 2012; Dembinsky, 2014) included remote locations. Twenty
studies were qualitative; four quantitative; two mixed methods; and one was a
systematic review (Kirby et al., 2016). Participants included patients with cancer and
non-cancer diagnoses. FCGs were either actively caring for a family member or
bereaved.
45
Table 2.4 Studies included in rural perspectives on end-of-life care literature review.
Author Year Country Method Diagnosis Informant Quality
Brazil et al. 2013, 2014 Canada Telephone survey Palliative Care FCG- active (44) and bereaved
(26)
High
Dale & Johnston 2011 Scotland Semi-structured ITVs Lung cancer Patients (6) High
Darer et al. 2015 USA Written survey Palliative Care NOK- bereaved (672) Medium
Dekker et al. 2012 South Africa Questionnaires, and ITVs HIV, TB, cancer, RF,
pneumonia
Patients (45); HCP (20) Low
Dembinsky 2014 Australia
(Indigenous)
Medical Anthropology-
informal ITVs, ethnography
Breast ca Patients (10), FCGs (4), non-
carer family (6), HCPs (5)
High
Devik/Wiik 2013/2011 Norway Phenomenology, narrative
ITVs
Cancer Older patients
(71-79years; n=5)
High
Devik et al. 2015 Norway Phenomenology, narrative
ITVs
Cancer Older patients
(71-92years; n=9)
High
Devik et al. 2016 Norway Semi-structured ITVs Cancer FCGs (bereaved; n=10) Medium
Duggleby et al. 2010, 2011 Canada Grounded theory; Open-ended
ITVs
Cancer Patients (6);
FCGs (bereaved; n=10)
High
Duggleby et al. 2014 Canada Written survey Cancer FCG (active; n=122) Medium
Grant et al. 2011 Uganda,
Kenya,
Malawi
Photographic ethnography,
ITVs, direct observation
Advanced illness Patients (33), FCGs (active
[17] & bereaved [10]), leaders
(29), staff (61).
Medium
Hansen et al. 2011 USA Semi-structured ITVs Cancer, chronic illness FCGs (bereaved; n=23) High
Hatcher et al. 2014 Australia 2 serial semi-structured ITVs Cancer FCG (active; n=6) Medium
Herce et al. 2014 Malawi Structured and open-ended
ITVs
Kaposi sarcoma, HIV,
cancer
Patients (36)
FCGs (active; n=11)
Included 9 dyads.
Medium
Howell et al. 2011 Canada Longitudinal; validated
assessment tools
Advanced disease. Patients (95) Low
Johnston et al. 2012 Scotland Serial ITVs Cancer Patients and their FCGs(n=20) Medium
Kelly et al. 2009 Canada Phenomenology, ITVs Not specified FCG (Aboriginal), bereaved
(10)
Medium
Kirby et al. 2016 Multinational Systematic review
Cancer and non-cancer 8 studies Medium
46
Table 2.4 continued
Author Year Country Method Diagnosis Informant Quality
Lockie/Pesut 2010/2010 Canada Semi-structured ITVs; written
questionnaire
Cancer Patients and FCGs (n=15) High
Mixer et al 2014 USA Semi-structured ITVs Life limiting illness Patients and/or FCG (active or
bereaved) n=11
Medium
Ostertag & Foreman 2008 USA FGD >50% cancer or HF FCG (bereaved, n=19), HCP
(53), volunteers (9)
Medium
Payne et al. 2007 UK Semi-structured ITVs Cancer or advanced
disease
Patients ≥65years (18) and
FCGs (11)
High
Pesut et al. 2011, 2014 Canada Ethnography, ITVs,
observation
Not specified Rural residents (95) included
FCGs (25)
High
Reece et al. 2014 USA ITVs Terminal illness Patient (1), FCG (10), hospice
care givers (9)
Low
Revier et al. 2012 USA Phenomenology; 2 interviews Not specified FCG (6) Medium
Tamannai et al. 2015 Cameroon Semi-structured ITVs Paediatric Burkett’s
lymphoma
Patients (3); FCG (7) High
Williams et al. 2013 Canada Participant daily journal;
narrative enquiry
Cancer FCG (active), n=23 Medium
Abbreviations: ITV- interview; FGD-focus group discussion; FCG-family caregiver; HCP-health care professional; HF-heart failure; TB-tuberculosis; RF-renal failure
47
2.6.1 Patients’ experiences and perspectives.
Rural patients identified several important features associated with maintaining quality
at EoL, and emphasised the importance of not giving up (Dale & Johnston, 2011; Devik
et al., 2013; Devik et al., 2015; Wiik, 2011); finding meaning in life (Devik et al., 2015),
while living life (Dale & Johnston, 2011); and maintaining dignity (Dale & Johnston,
2011; Devik et al., 2013; Grant et al., 2011; Johnston et al., 2012), independence (Dale
& Johnston, 2011; Devik et al., 2013; Devik et al., 2015; Duggleby et al., 2010, 2011,
2014; Johnston et al., 2012) and normality (Dale & Johnston, 2011; Devik et al., 2013;
Johnston et al., 2012). This required most patients to redefine ‘normal’ (Dale &
Johnston, 2011; Devik et al., 2013; Duggleby et al., 2010; Johnston et al., 2012; Wiik,
2011), come to terms with change (Devik et al., 2015; Duggleby et al., 2010; Johnston
et al., 2012), [often with resignation (Dale & Johnston, 2011) or struggle (Devik et al.,
2013)], while making the most of every day (Dale & Johnston, 2011). Patients
described how they lived with exhaustion and stress (Devik et al., 2013; Duggleby et
al., 2010, 2011; Wiik, 2011) often while balancing hope and despair (Devik et al.,
2013). While a patient’s deteriorating physical condition increased their dependency on
others, most continued to fiercely defend their independence (Duggleby et al., 2011;
Johnston et al., 2012) for as long as was possible. Some patients talked openly about
impending death, while holding onto hopes and dreams for the future (Devik et al.,
2013; Johnston et al., 2012; Wiik, 2011); while others described enduring life bravely,
despite being too exhausted to enjoy life (Devik et al., 2013; Wiik, 2011).
2.6.1.1 Maintaining dignity
Patients maintain dignity by refusing to be defined by their illness (Dale & Johnston,
2011; Devik et al., 2013; Duggleby et al., 2010, 2011; Johnston et al., 2012) and
searching for hope (Devik et al., 2013, 2015; Duggleby et al., 2010; Grant et al., 2011;
Johnston et al., 2012) as ‘hope is the key to enduring distress’ (Devik et al., 2013,
p. 785). Many patients were reluctant to ask for help, as they were afraid of becoming a
burden (Dale & Johnston, 2011; Devik et al., 2013; Duggleby et al., 2010, 2011;
Johnston et al., 2012) and/or losing independence (Devik et al., 2013; Duggleby et al.,
2011). One Norwegian patient found hope in pursing life-prolonging chemotherapy
despite losing dignity through side-effects (Devik et al., 2013), while others refused
chemotherapy to maintain their quality of life (Johnston et al., 2012). None of the
participants spoke directly of euthanasia.
48
2.6.1.2 Preparing for death
Preparing for death, such as funeral planning, ‘without loss of hope or the desire to keep
living’ was important for some patients. (Johnston et al., 2012, p. 1622). Some patients
found comfort in preparing themselves and their families for impending separation
(Dale & Johnston, 2011). This was achieved through activities such as funeral planning
and completion of wills.
2.6.1.3 Pain control
Most participating dying people and families feared pain. In sub-Saharan Africa, pain
dominated the lives of patients (Grant et al., 2011; Herce et al., 2014). In one Malawi
study, 86% of participants reported pain to be moderate to severe (Herce et al., 2014).
Inadequate pain control was reported in four studies from developed countries (Devik et
al., 2015; Johnston et al., 2012; Pesut et al., 2010; Wiik, 2011). Participating patients
described pain management as important (Devik et al., 2015) and essential for self-care
(Johnston et al., 2012). Anticipating medication requirements, when commuting long
distances, was an important consideration for some patients (Pesut et al., 2010; Lockie
et al., 2010).
2.6.1.4 Formal and informal support
Support of family, friends, community (Duggleby et al., 2011; Herce et al., 2014) and
HCPs (Duggleby et al., 2010) was considered essential, and valued. Despite this support
and connection, the message from two studies was that patients felt isolated as their
disease progressed, ‘I’m part of the community but I feel alone’ (Duggleby et al., 2011,
p.2) and having to ‘walk the palliative path alone’ (Wiik, 2011, p. 12).
2.6.2 Family caregivers’ experiences and perspectives
To obtain a broad view of rural EoLC it was important to hear the voice of the rural
FCGs, and to explore not just the common views, but to also determine the degree of
congruence between perspectives. Family caregivers often felt responsible for ensuring
the dignity and comfort of their loved ones during their final days, and spoke of their
distress, guilt and anger when circumstances prevented them fulfilling this role (Revier
et al., 2012). While rural FCGs took on the responsibility of providing direct care, many
also took on the role of managing and coordinating care, while being an advocate for
49
their family member (Pesut et al., 2014); however, ‘few [were] physically, emotionally,
or educationally prepared for the tasks and responsibilities of caregiving’ (Revier et al.,
2012, p. 5), especially as the illness progressed (Duggleby et al., 2011).
FCGs experienced a broad spectrum of negative emotions (Williams et al., 2013). If
FCGs could find meaning in the situation, by focusing hope on the day-to-day moments
(Revier et al., 2012), redefining normal (Duggleby et al., 2010; Williams et al., 2013),
connecting and separating (Revier et al., 2012), they were better positioned to cope with
the responsibilities, burden of care and transitions in roles (Duggleby et al., 2010). Self-
care and maintaining their own quality of life was important; however, Williams et al.,
(2013, p. 6) reported that FCGs often lacked the ‘time and energy to do everything they
needed and to care for themselves.’ To fulfil their caregiver responsibilities, FCGs
required support (Duggleby et al., 2011; Pesut et al., 2011, 2014; Brazil et al., 2013,
2014; Herce et al., 2014) from family, friends, neighbours and HCPs and they
‘identified the need of having someone to talk to, and being appreciated by the care
recipient’ (Brazil et al., 2014, p. 15).
Some FCGs accepted the role out of obligation (Herce et al., 2014); however, many
found meaning in caring for their loved ones (Revier et al., 2012; Pesut et al., 2014),
with the burden of care ‘outweighed by … the satisfaction they derived from having
made a meaningful contribution’ (Pesut et al., 2014, p. 130). However, many FCGs
become fatigued with full time care, and with the responsibility of providing care to a
loved one.
2.6.3 Patient and FCGs’ common experiences and perspectives
2.6.3.1 Communication; accurate and timely information
The greatest support need of participating rural patients and FCGs, in developed
countries, was informational. Effective communication between HCPs and
patients/FCGs, and within families (Tamannai et al., 2015), reduced pain and distress
(Grant et al., 2011), empowered carers to fulfil their responsibilities (Revier et al.,
2012), facilitated smooth transitions of care (Hatcher et al., 2014), and allowed patients
and families to prepare for death (Johnston et al., 2012; Dembinsky, 2014). Most, but
not all (Devik et al., 2013) participants were satisfied with the standard of
communication by rural HCPs. However, many participants reported one or more
50
communication difficulties, such as receiving conflicting or untimely information
(Ostertag & Foreman, 2008; Pesut et al., 2014), uncertainty as to ‘which physician was
in charge’ (Darer et al., 2015), and not receiving information from the person
considered to be the expert (Duggleby et al., 2010; Pesut et al., 2014). While some
patients accepted medical advice without questioning (Dale & Johnston, 2011), others
considered false hope (Kelly et al., 2009) or poor communication, a lack of respect for
the patient and their family (Kelly et al., 2009; Duggleby et al., 2010, 2011) and that
‘not knowing was worse than knowing’ (Johnston et al., 2012, p. 1623). Ostertag and
Foreman (2008) reported primary care physicians were highly praised for honesty and
presence at the time of death.
2.6.3.2 Formal services
Brazil et al. (2014) reported 82.6% of their participating rural FCGs indicated that
formal palliative care services were readily available in the rural locations where they
resided, with 68.6% having access to services after hours. While the participants in the
study by Pesut et al. (2011) reported that a HCP was always available, these results
were not universal. Other rural studies reported inadequate accessibility to care and
continuity of care (Duggleby et al., 2011; Wiik, 2011; Hansen et al., 2012; Johnston et
al., 2012; Devik et al., 2013), especially after hours (Duggleby et al., 2010, 2011;
Hansen et al., 2012; Johnston et al., 2012). Access to HCPs with palliative care training
(Duggleby et al., 2010), paid qualified in-home carers (Hansen et al., 2012), after-hours
pharmacies or morphine (Grant et al., 2011; Hansen et al., 2012), respite care (Brazil et
al., 2013, 2014), and paediatric hospice (Mixer et al., 2014), are often limited or
unavailable in rural areas.
Features of care that facilitated quality rural EoLC included personalised care
(Duggleby et al., 2010; Hansen et al., 2012); ‘knowing’, and ‘being known’ by the
HCPs, and a willingness of HCPs to go beyond their professional care (Pesut et al.,
2011). However, loss of privacy and anonymity, and an expectation that friends will
always be available were perceived as barriers (Pesut et al., 2011). Some participants
reported the quality of care provided was dependent on the personality of the HCP, with
difficulties arising if personality conflicts arose, as often no alternative provider was
available (Devik et al., 2013).
51
When care at home was not possible the local community hospital was an acceptable
alternative (Hansen et al., 2012; Herce et al., 2014; Mixer et al., 2014). Rural hospitals
were considered safe (Hansen et al., 2012), small, convenient, personal, and welcoming.
Rural nursing staff were described as caring and compassionate. However, rural
hospitals were not viewed positively by indigenous participants in Australia
(Dembinsky, 2014) and Canada (Kelly et al., 2009), mainly due to cultural insensitivity.
2.6.3.3 Informal social support - family, friends and neighbours
Nineteen rural studies reported on the informal support provided by family, friends,
neighbours and the community, with some participants stating that family is the ‘most
important’ factor (Devik et al., 2013, 2015) and essential for culturally congruent care
(Mixer et al., 2014). Brazil et al. (2013, 2014) reported that the greatest unmet needs
identified by rural FCGs were the tangible or practical needs. Community support was
reported to have a positive influence on rural EoLC with one Norwegian participant
describing a sense of solidarity, as people took care of each other (Devik et al., 2015).
However, not all participants admitted to having happy family relationships (Mixer et
al., 2014; Williams et al., 2013). Other rural participants acknowledged that community
support cannot be taken for granted and is highly reciprocal. Those participants who had
been involved in giving to their rural community often received the highest amount of
support from that community (Pesut et al., 2011). Despite the strong sense of
community, studies reported that as disease progressed and patients lost mobility and
independence there was a sense of isolation, with one participant saying, ‘I am part of
my community but I feel alone. Family and friends come to visit me, but I feel isolated
as they are unable to understand what is happening to me and my wife’ (Duggleby et
al., 2011, p. 2).
2.6.3.4 Emotional support
Strong emotional support was identified as a facilitator of quality rural EoLC and was
dependent on good communication, information, the presence of HCPs, support of other
patients (Johnston et al., 2012), faith and hope. Hope was maintained through
connection with family (Devik et al., 2013), friends, and being linked to something
outside the illness (Revier et al., 2012).
52
2.6.3.5 Spiritual support
Spiritual connection and faith fostered hope (Kelly et al., 2009; Revier et al., 2012;
Williams et al., 2013; Mixer et al., 2014) with faith seen as an enabler to persevering in
life as death drew near. Faith was reported to be fundamental to rural Appalachians and
their transition through EoLC (Mixer et al., 2014). In many rural communities, church
support was not limited to spiritual issues, as congregations also provided physical and
financial support (Tamannai et al., 2015).
2.6.4 Influence of rural place and culture
Despite diversity in rural settings, rural residency was seen as having positive and
negative influences on rural EoLC for both patients and FCGs. Distance was identified
as the greatest negative influence. For many participants, commuting for treatment was
considered stressful and exhausting (Devik et al., 2013; Duggleby et al., 2011; Lockie et
al., 2010); inconvenient (Devik et al., 2013) and expensive (Grant et al., 2011; Hansen
et al., 2012; Lockie et al., 2010; Pesut et al., 2010). Travelling impacted negatively on
the health of some FCGs (Lockie et al., 2010) and resulted in fragmented care
(Duggleby et al., 2011). Some rural participants accepted commuting (Devik et al.,
2015) as ‘one of the compromises they have to make for living at home, that is, to live
in a place that contributed to their overall health’ (Wiik, 2011, p. 12).
Geographical distance limited accessibility to home-based services as some patients
lived outside the boundary for home visits (Hansen et al., 2012), visits were less
frequent especially in bad weather (Lockie et al., 2010; Pesut et al., 2010; Dembinsky,
2014) and were often not available at short notice or after hours (Duggleby et al., 2010,
2011; Hansen et al., 2012). However, opinions regarding the effect of distance on the
quality of care were divided with some seeing it as a major barrier to rural residents
receiving home-based palliative care and therefore achieving a home death (Dembinsky,
2014), and others not viewing distance and rural residency a disadvantage (Devik et al.,
2015). However, with advanced illness, the participants’ sense of solitude became one
of isolation (Duggleby et al., 2011). Geographic isolation also explained the greater
unmet emotional needs of rural caregivers as they usually lacked the support of others
going through the same experiences (Brazil et al., 2014).
53
Mobile phones (Grant et al., 2011), computers and internet access (Devik et al., 2015;
Pesut et al., 2010) helped reduce the sense of rural isolation by maintaining contact with
distant family and improving access to HCPs; however, these technologies are not
available everywhere (Lockie et al., 2010; Pesut et al., 2010). For some patients, a
phone call was not sufficient, and did not replace the physical presence of HCPs (Pesut
et al., 2010; Hansen et al., 2012).
2.6.5 Meaning of home or ‘home country’
In one study, 46% of patients and FCGs reported their preferred place of care to be
“home” (Herce et al., 2014), as ‘being at home is like a brick being in the right place:
this is my land and these are my people’ (Devik et al., 2015, p. 7). Many of the rural
participants were entrenched in their community and had a strong place attachment
(physical, social, and autobiographical). Devik et al. (2015, p. 8) suggest the rural
environment had ‘potential to be a source of comfort, security, and identity’, while
others reported the urban environment was often viewed negatively (e.g. noisy,
unfamiliar, and unknown spaces).
For indigenous rural residents, home or ‘home country’ has special cultural
significance. The biggest barrier to using hospital-based palliative care services for these
participants was not being able to die ‘in country’. The lack of cultural awareness by
HCPs and misperceptions of the concept of palliative care were barriers to accepting
palliative care (Grant et al., 2011; Dembinsky, 2014).
2.6.6 Summary of rural end-of-life care from the experiences and perspectives of
patients and family caregivers
Consistent with previous reports (Hughes et al., 2004), rural participants in this
literature review were mostly satisfied with the EoLC provided to them; however, most
were realistic and openly acknowledged their unmet needs. The over-riding themes for
patients and FCGs in all locations were ‘living life’; holding onto hope, dignity and
meaning; receiving personalised care; being known; and the desire for HCPs to be
present, provide reassurance, and honour the choices of patients and FCGs. In sub-
Saharan Africa, this was possible once pain was managed. The EoL issues faced in
general by patients and FCGs, regardless of where they live (Lynch et al., 2013;
Ciemens et al., 2015; Sandsdalen et al., 2015), were raised by the rural participants in
54
this review; however, there are differences, facilitators and barriers (such as the
hardship of distance), unique to rural settings that significantly impact rural EoLC.
Quality EoLC facilitates a ‘good death’ in rural settings.
While there is no expectation that resources in rural areas should be equivalent to those
available in urban settings (Kaasalainen et al., 2014), some rural HCPs lamented their
lack of palliative skills, training and mentoring (Castleden et al., 2010; Goodridge &
Duggleby, 2010b). This insufficiency is especially significant in the hospital setting.
Rural hospitals often act as a substitute for inpatient hospices (Wilson et al., 2006; Spice
et al., 2012), and so EoL services must be integrated into the healthcare provided. (Fink
et al., 2013).
2.7 Australian rural studies
While the above review details rural EoLC across heterogeneous rural settings, I had
intended to close my literature review by focusing on studies conducted in similar rural
settings to that of my thesis, or as a minimum, to report on relevant Australian studies.
However, there is a paucity of Australian rural studies specifically focussing on a ‘good
rural death’ or preferred place of care and death. The Australian studies have mostly
been related to indigenous populations (Diaz et al., 2015; McGrath, 2007) or have
included urban and rural residents without sufficiently identifying the rural views
(Kirby et al., 2016; Pereira, 2005; O’Connor & Lee-Steere, 2006; Buikstra et al., 2006)
2.8 Chapter summary
In this chapter, I have outlined the current rural research literature relating to the rural
‘good death’, end-of -life care, and place of death. The studies identified are a
combination of qualitative and quantitative research. Where possible I have reported the
findings from the experiences and perspectives of rural patients, with a life limiting
illness, and their family caregivers.
The chapter started by defining the discipline of palliative care (Section 2.1). The
origins and development of palliative care were explored. One of the aims of
55
contemporary palliative care is to assist people receiving EoLC to die at home, if that is
their preference. This requires the support of family and a multidisciplinary team of
formal and informal providers. Some studies have suggested that a home death implies
a ‘good death.’
Section 2.2 described the evolution of the concept of the good death. The concept is
dynamic, having changed from the dying person having an awareness of impending
death, to the western contemporary ‘controlled’ death. The ‘good death’ is open to
interpretation based on personal, social, cultural, political and religious perspectives.
Formerly a very public affair, the contemporary ‘good death’ is often hidden away in
institutions, such as hospitals and RACFs.
The definition of ‘rural’ was explored in Section 2.3. Currently, there is no consensus
on the definition of ‘rural’ making it difficult to compare rural studies and their
findings. The current rural studies are heterogeneous in terms of location, population
and resources. However, rural residents describe themselves as different to urban
residents. There are unique factors that either hinder or facilitate the rural ’good death’
and determine rural residents’ actual PoD.
Due to the paucity of rural studies, one scoping and two systematic literature reviews
were undertaken. The findings of these reviews are reported in this chapter. The
findings of ‘Rural residents’ perspectives on the rural ‘good death’: a scoping review’
are reported in Section 2.4. The ‘good death’ referred to the death event, the dying
process, the meaning of death and the after-death concept. The elements considered
essential for a ‘good death’ were those that addressed the physical, emotional, social,
spiritual and cultural needs of the dying person and their FCGs. Alternative names for
the ‘good death’ included controlled, timely, dignified, social and noble death. There
were rural factors that either hindered or facilitated a ‘good death’. Place of death, and
in particular, dying at home, was one factor that facilitated a ‘good death’.
Place of death in rural settings was further explored by systematically reviewing the
literature. Section 2.5 reported the findings of ‘Place of death in rural palliative care: a
systematic review’. This review reported on the actual PoD of patients receiving
palliative care in rural settings, and the preferred PoD, as expressed by rural patients
56
and/or their FCGs. Most deaths in rural settings occurred away from home, with most
occurring in hospital. The determinants of PoD were similar to those identified in urban
locations: patient functional state and clinical condition, carer capacity, social networks
and health system facilities and resources. Most rural residents were less likely to die at
home that urban residents, and were disadvantaged by rural residency and lack of
options. Rural hospitals were considered a substitute for inpatient hospice. Some studies
reported PoC and PoD were not necessarily the same, and that people change their
preferences along the disease trajectory. Rural residents indicated that PoD was not a
major concern of the dying and providing they died within community, home was not
essential for a ‘good death’.
The findings of the second systematic review, ‘Rural end-of-life care from the
experiences and perspectives of patients and family caregivers: a systematic literature
review’ were reported in Section 2.6. Limited studies were identified. Features
considered important for quality EoLC included maintaining dignity, hope,
independence and normality; the presence of family and friends; minimising carer
burden; timely and accurate information; formal and informal support; physical,
emotional and spiritual support; living life, personalised care, and being known.
Residents in developing countries had different priorities to those in developed regions,
with access to adequate pain control being the greatest unmet need.
This comprehensive and systematic review of the literature has highlighted a significant
gap in the current rural literature. Presently, there is a lack of rural end-of-life studies
reporting specifically on the ‘good death’ and PoD experiences and perspectives of rural
patients, with a life limiting illness, and their family caregivers. Where experiences and
perspectives have been reported, they have mostly related to patients with cancer or
HIV/AIDs. The voice of rural patients, dying of advanced chronic non-cancer diseases,
are mostly unheard. Therefore, there is insufficient evidence to draw any reliable
conclusions on what facilitates or hinders a rural ‘good death’, or the importance of
home as a PoD.
57
CHAPTER THREE: THE SNOWY MONARO –
DESCRIBING THE SETTING AND POPULATION
He hails from Snowy River, up by Kosciusko's side,
Where the hills are twice as steep and twice as rough,
Where a horse's hoofs strike firelight from the flint stones every stride,
The man that holds his own is good enough.
And the Snowy River riders on the mountains make their home,
Where the river runs those giant hills between;
I have seen full many horsemen since I first commenced to roam,
But nowhere yet such horsemen have I seen.
(THE MAN FROM SNOWY RIVER by A.B. "Banjo" Paterson, 1890.)
The literature review supports the argument that each rural community is unique.
Therefore, it is appropriate to allocate a full chapter to the setting of this research study.
A visual image of the setting, created for the reader, will assist in placing the findings of
this study, into a meaningful context. This research study is conducted within the
Snowy Monaro region of NSW, Australia. Section 3.1 outlines the reasons for choosing
this region, and briefly places the Snowy Monaro within the Australian context. This is
followed, in Section 3.2, by describing this rural setting – the location, demographics,
history, medical services, transport, and community services. Section 3.3 briefly
introduces the target population. Section 3.4 discusses the challenges faced in recruiting
participants to end-of-life research. The chapter concludes (Section 3.5) with a summary
of the chapter.
3.1 Introduction
Being a rural resident and rural medical practitioner, the chosen setting was the location
in which I am familiar, the Cooma Monaro and Snowy River Shires (Local Government
Areas). To gain a cross-section of views, including those from more remote areas with
reduced access to medical resources, the neighbouring Bombala Shire was also
included. At the time this seemed appropriate as all three Shires made up the Monaro
Health Service, a sub-division of the Southern New South Wales Local Health District
(SNSW LHD). As chance would have it, the NSW Government enforced a compulsory
58
amalgamation of local government councils, and on 12th May 2016, at the end of my
data collection period, the three Shires of Cooma-Monaro, Bombala and Snowy River,
merged to become the Snowy Monaro Regional Council. This is a diverse region
comprising towns, villages and farms with residents engaged in agriculture, forestry,
hydroelectric production, tourism (skiing, fishing, mountain biking), National Parks,
and a variety of support industries (education, health, policing, local government,
tradesmen, hospitality, and retail).
To enable the reader to place this study into the Australian context, a brief description of
the Australian population is provided. This is followed by a more focused description of
the region under study. The Australian population, as of 18th June 2017, was 24.5
million (ABS ‘Population clock’, 2017b), with approximately two-thirds of the
population living in a capital city. At June 30, 2016, the estimated resident population of
NSW was 7.7 million people, with just over one third of NSW residents living outside
Greater Sydney (Table 3.1).
Table 3.1 Estimated resident population (ERP) by remoteness structure. (ABS Regional
population growth, 2017d)
3.2 The Snowy Monaro region
The area chosen for the study, the three previous NSW local government areas (LGA)
of Snowy Monaro, Bombala and Snowy River Shires, located in south eastern NSW
(Figure 3.1), covers an area of 15,162 km2 with a population of 20,218 (ABS Census
ERP at 30 June CHANGE
2015
No.
2016
No.
2015-2016
No. %
NSW
Major Cities 5675992 5765757 89 765 1.6
Inner Regional 1458267 1473700 15433 1.1
Outer Regional 448098 448862 764 0.2
Remote 30638 30392 -246 -0.8
Very Remote 8344 8231 -131 -1.6
Total 7621339 7726924 105585 1.4
Australia
Major Cities 16870359 17159014 288655 1.7
Inner Regional 4308519 4357568 49049 1.1
Outer Regional 2086950 2090572 3622 0.2
Remote 321194 319314 -1880 -0.6
Very Remote 204033 202408 -1625 -0.8
Total 23791055 24128876 337821 1.4
59
QuickStats, 2017a). The demographic features are reported in Table 3.2. This region is
serviced by the Monaro Regional Health Service. This region borders the Australian
Capital Territory, and the Queanbeyan-Palerang LGA to the north; the Bega Valley to
the east; the Snowy Mountains to the west; and Victoria to the south (Figure 3.2).
Cooma is the regional centre and largest town, with a population of 6742 (August
2016). Smaller towns include Bombala (population 1386), Jindabyne (2629), and the
villages of Berridale (1197), Thredbo (471), Bredbo (352), Delegate (351), Adaminaby
(301) and Nimmitabel (320) (ABS Census QuickStats, 2017a). There are also smaller
communities and many farms.
Figure 3.1: Map of NSW with study location indicated by arrow.
The population density ranges from 0.6-2 /km2, with an average population density of
1.4 residents per km2. The population in the south-western area (Jindabyne and
surrounds) increases by approximately 10,000 (Monaro Regional Health Services Plan,
2015) in winter, due to the influx of seasonal workers and tourists during the ski season.
This region was chosen for its diverse rural characteristics, absence of a specialist
multidisciplinary palliative care service, and my familiarity with the region. Public
health services are managed by the Monaro Regional Health Services (MRHS),
governed by the Southern NSW Local Health District (SNSW LHD).
Study area
60
Monaro Regional Health Services (MRHS, 2015, p. 10).
A significant proportion of the Monaro population do not live in the main towns.
Approximately one third of Cooma-Monaro LGA residents live outside Cooma, two
thirds of Snowy River LGA residents live outside Jindabyne, and half the Bombala
LGA population lives outside the Bombala Township (MRHS, 2015).
Table 3.2: Demographic features of the Snowy Monaro residents.
(ABS QuickStats, 2017a)
Snowy Monaro NSW Australia
Males 51.7% 49.3% 49.3%
Median age (years) 43 38 38
>65 years 19.5% 16.2% 15.8%
Median weekly
household income
$1200 $1486 $1438
Aboriginal and/or
Torres Strait Is. people
2.2%
2.9% 2.8%
Both parents born in
Australia
63.2% 45.4% 47.3%
3.2.1 Snowy Monaro history
The rural culture evident today in the Monaro region has been moulded by its history.
The Cooma-Monaro Shire land is an ancient plateau running north south about 800
metres above sea level, and bounded on the east and west by rugged mountain ranges,
Figure 3.2: Map of the Snowy Monaro region
61
part of the Great Dividing Range. The name ‘Monaro’ comes from the aboriginal word
‘Monaroo’, meaning high plateau or plain. The tablelands are naturally treeless due to
soil types, low rainfall, low temperatures and cold air pooling (Figure 3.3).
Figure 3.3: Monaro Plains, outside Cooma (2016)
Photograph: Suzanne Rainsford
The south-western corner of the Snowy Monaro region is mountainous, and while the
elevation is much lower than the mountains in Europe, North America and central Asia,
the area is snow covered during the winter months (Figure 3.4).
Figure 3.4: The Main Range, Kosciuszko National Park
Photograph: Suzanne Rainsford
62
Archaeologists believe that Aboriginal people have been living in the Snowy Monaro
region for as long as 20000 years before the arrival of Europeans. The two main tribal
groups were the Ngarigo people of the tablelands, and the Wogul or Wolgalu people of
the high country. During bad weather it is believed the tribes moved to the warmer
valleys. During the warmer months, large-scale inter-tribal gatherings were held in the
high country, when tribes travelled from the South Coast to collect the large bogong
moths (Plowman et al., 1999, p. 1). Following European settlement, in the 1800’s, the
number of Aboriginal people declined. Records during the 1840s reported a group of
between 500 and 1,382 Ngarigo were frequently seen on the Monaro, however, by 1856
only 166 Aboriginal people were counted in the census, and by 1892 just two remained.
It was originally believed the Ngarigo disappeared due to disease and interactions with
European settlers, but recent information suggests many moved to the coast and
integrated with the coastal clans around Bega and Bermagui (Plowman, 2007). The
2016 census identified 448 Snowy Monaro residents as Aboriginal or Torres Strait
Islander, representing 2.2% of the population (ABS Census QuickStats, 2017a). While
this study did not specifically include or exclude indigenous participants, this
demographic was not collected.
Captain Mark Currie and Major John Ovens carried out the first expedition to the
Monaro, arriving on 4 June 1823. It was not long before squatters, shepherds and
herdsmen arrived, bringing with them flocks of sheep and herds of cattle belonging to
wealthy Sydney landowners. Today, the Monaro is known for its high-quality Merino
wool and beef cattle (Figure 3.5). In 1833, the first squatters arrived in the Bombala area
and took ownership of what was to become the Bombala Station. The Snowy Mountain
area was first explored by Europeans in 1835. Edmund Strzelecki, ascended the highest
peak, Mount Kosciuszko, in 1840 and named it after a Polish patriot. High-country
stock men followed, using the Snowy Mountains for grazing during the summer
months. The town of Cooma dates back to 1842, when John Lambie, Crown Lands
Commissioner for Maneroo, constructed a residence and office (Plowman, 2007).
63
Figure 3.5: Cattle drive on the Snowy Mountains Highway
Photograph: Suzanne Rainsford
The discovery of gold in 1859, on the high plains near Kiandra, boosted Cooma’s
development. At its height, Kiandra had a population of about 4,000 people, and ran 14
hotels. The gold rush was short lived, ending in the mid 1860’s. Kiandra was
abandoned, resulting in an influx of people looking for work in Cooma. The town
continued to prosper, and by the second half of the 19th century Cooma became the
commercial centre for the Monaro.
In 1949, Cooma was chosen as the headquarters for the Snowy Mountain Hydro-electric
Scheme, the construction of a world class hydroelectric system supplying electricity and
water for irrigation. This quiet rural town was rapidly transformed into a major centre,
with the population increasing from 2,000 to 10,000 people. Other new towns sprang up
in the surrounding Snowy Mountains. Cooma was the first multi-cultural town in
Australia, with two thirds of the Snowy workers being immigrants from war torn
Europe, representing 40 countries. The Scheme brought prosperity to the town and
region, and confirmed Cooma as the commercial centre of the Monaro.
Following the completion of the Snowy Hydro Scheme, while many workers drifted
away, a significant number remained in the region, living in towns, villages and farms.
The ‘Snowy’ people were hardy. I can remember, a few years ago, one elderly German
patient recounting stories of her early years in the area. The family (husband and two
64
small children) arrived from war torn Germany and spent the first 18 months, including
two snowy winters, living in a tent, with no electricity, just out of Jindabyne.
Today, Cooma is a busy commercial and tourist centre and the ‘Gateway Capital to the
Snowy Mountains’ (Figure 3.6). It is located at the junction of two highways linking
Canberra (115 km), the south coast (115 km) and the mountains (100 km). Over the
years, the region has become known for its Merino sheep, Angus and Hereford cattle,
farming, forestry, and as a popular tourism destination (skiing, fishing, and more
recently mountain bike riding). The average resident endures drought, bushfires,
summer heat, winter snow and ice, fog, rabbits, foxes, weeds, wool prices up, wool
prices down, and many other adversities. Yet this is their home, and for many this is
where they want to live and where they choose to die.
Figure 3.6: Cooma; population 6742 (August 2016)
Photograph: Suzanne Rainsford
3.2.2 Medical facilities
The Government Gazette of 1844 records two doctors officially listed for the Monaroo
Plains. One held a stock run in the Bombala district. A prerequisite for a Monaro doctor
was that he was ‘a strong and capable rider, to be available at any time day or night and
be prepared to ride many miles in an emergency’ (Plowman, 2007 p36).
Cooma’s first permanent doctor arrived in the village of 166 people in 1865. A small
hospital, built with government and private funds, was opened in 1867. Today, Cooma
has a 37-bed public hospital (30 medical/surgical beds and 7 obstetric beds). The
hospital has a busy emergency department, day surgery/endoscopy unit, inpatient
operating theatre, maternity ward, outpatient oncology and renal dialysis units,
physiotherapy (with hydrotherapy pool), pathology and radiology department including
65
CT scanning. In addition, there is a multidisciplinary community health service
including one palliative care nurse (0.9 full time equivalent [FTE]) and one McGrath
Breast Care Nurse1 (0.3 FTE) who doubles as a community nurse. Medical cover is
provided by local general practitioners. In addition, there are visiting specialists in
general surgery, orthopaedics, gynaecology, endoscopy, gerontology and psychiatry.
While not having formally designated palliative care beds, two single rooms have been
set aside for patients receiving end-of-life care. The former staff meeting room has been
beautifully refurbished and renamed the Mary Green Room2, in memory of Cooma’s
first oncology nurse who died a few years ago from cancer. When required, this lounge
room provides private day and overnight accommodation for families of dying patients.
The hospital has on an average 14.3 palliative care deaths/year (range 10-19) from July
2008 to June 2013 (MRHS, 2015).
The Snowy Monaro region has eight General Practice surgeries. Surgeries in Cooma
and Jindabyne are open Monday to Friday; practices in smaller towns are part-time.
After hours’ services are provided by the emergency departments at Cooma and
Bombala hospitals. Cooma has two large general practices staffed by 14 FTE GPs. I
currently work in Cooma (0.2 FTE) in general practice and as a private palliative
medicine specialist. There is no specialist multidisciplinary palliative care team or local
inpatient hospice. Most of the end-of-life care is primary palliative care. Cooma has two
residential aged care facilities (RACF) (72 beds and 40 beds) offering respite and
permanent high level care placement.
Bombala has a multi-purpose service (MPS) with an inpatient facility (seven medical,
one palliative care and ten RACF beds), community health services and a part-time
general practice. The Bombala MPS has limited medical support, especially after hours.
Most acute or serious injuries or illness are transferred to either Cooma Hospital, or the
Bega Regional Hospital. Bombala also has a privately run RACF with 32 beds.
1 McGrath Breast Care Nurses are sponsored by the McGrath Foundation, to help individuals (and their
families) experiencing breast cancer by providing physical, psychological and emotional support. This
support is available for free, from the time of diagnosis, and throughout treatment and end-of-life.
2 Mary Green Memorial Palliative Care Family Room. Opened in 2015, on the main ward of Cooma
Hospital, adjacent to the two single rooms set aside for EoLC. Funded by MCCR, in honour of Mary
Green, the first oncology unit nurse. In 2012, she lost her battle with cancer. Facilities include a sofa bed,
TV, equipped kitchenette and dining suite. This room is for the use of families of palliative care patients.
66
Delegate, 35 kms south of Bombala, towards the Victorian border (Figure 3.7), has a
MPS with a nine-bed RACF. The emergency department is nurse led and offers basic
first aid. The palliative care nurse, based in Cooma, visits Bombala weekly and
Delegate weekly, when required.
Figure 3.7: Delegate; population 351 (August 2016)
Photograph: Suzanne Rainsford
Despite strong campaigning by the locals, currently there are no inpatient facilities or
after hours’ medical services, in Jindabyne. The general practitioners conduct their
practice from two private medical practices. Community health services, from Cooma,
are available in a purpose-built facility, including palliative care (nurse) home visits.
The closest inpatient hospital is Cooma (1-hour drive in good weather). There is a 14
bed RACF in Berridale (half way between Cooma and Jindabyne). As previously
mentioned, the population of Jindabyne swells by 10,000 during the winter months.
Travel times in winter increase due to ski traffic, fog and icy roads.
Other medical practices in the Snowy Monaro region are located in the winter ski
resorts of Thredbo and Perisher Valley (Figure 3.8). The medical centres in Thredbo
and Perisher are open during the winter season. Thredbo medical centre is also open
during Easter and summer school holidays.
67
Figure 3.8: Sharp Street Surgery, Cooma and Perisher Valley Medical Centre.
(Photographs: Suzanne Rainsford)
The palliative care and community nurses are available to visit all homes within the
region; however, the degree of isolation dictates the frequency of visits. While some
oncology patients receive their chemotherapy in Cooma, there are no visiting
oncologists. Most patients requiring specialist medical care travel to Canberra, although
a small but significant number travel to Bega (114 km from Cooma) or even further
away to Sydney (450 kms) or Melbourne (620 kms). Patients requiring tertiary medical
care are transported to The Canberra Hospital, either by road or retrieval helicopter.
Most of the palliative care within the region is provided jointly by the general
practitioners supported by the palliative care Clinical Nurse Consultant (CNC). While
there is a multidisciplinary community health service, currently there is no counsellor,
social worker or palliative care pastoral care worker. After completing my training and
gaining Fellowship of the Australasian Chapter of Palliative Medicine (FAChPM) I now
offer a part-time (0.2 FTE) private palliative medicine consultancy service in Cooma.
The closest specialist multi-disciplinary team and specialist palliative care inpatient
hospice are in Canberra (115 kms). New state and commonwealth government
initiatives were implemented after completion of my data collection. There is now a
HammondCare in-home package3 available to assist patients and their FCG in the final
days of life (MacLeod et al., 2015); an after-hours telephone palliative care advisory
3 HammondCare Palliative Care Home Support Program - provides free, non-means tested end-of-life
care for patients who wish to remain at home. A package of up to 48 hours of specialised supportive
palliative home-based care, day or night, is provided by specially trained community care workers who
work as part of the existing multidisciplinary team. http://www.hammond.com.au/services/palliative
68
service; and an ambulance after-hours palliative care plan. These initiatives have the
potential to improve the quality of palliative care in rural areas, especially after hours.
The telephone advisory service is available to patients, families, and healthcare
providers, both in community and hospital settings. A systematic literature review by
Mitchell (2002) reported the majority of palliative care in the community is provided by
GPs; however, while patients and FCGs were appreciative of the care provided, GPs
often felt inadequate and unsupported. The telephone advisory service is one step
towards providing specialist support in rural areas.
3.2.3 Transport
As in any rural region, distances and transport compound the issues of isolation. In
Cooma, the train line from Sydney opened in 1889, but closed in 1989. The train has
been replaced by a bus, the Transborder Express Trainlink. This bus travels by road
from Bombala to the Canberra railway station (four hours), from where a train line runs
directly to Sydney (four hours). The bus travels from Bombala (and return) three days a
week, via Jindabyne and Cooma, to Canberra train station, where taxis are available for
transfer to the Canberra airport. There is also a daily bus service from Bega via Cooma
to Canberra. Cooma currently has a small airport with one flight to Sydney, Monday to
Friday. Additional flights are available during the ski season. Due to the frequent fog,
flights are often delayed. Unless a small town lies directly on the bus route, the only
available public transport is the local school bus. Residents in Adaminaby needing to
use public transport can travel to Cooma, for the day, by school bus.
The additional medical costs of travel and accommodation are partly offset by a
government funded Isolated Patient Transport and Accommodation Scheme. This
provides a small reimbursement of travelling costs, providing the patient travels over
100 kms to the nearest medical service. In addition, for a small fee, patients can use the
subsidised Health and Community Transport Scheme to attend medical appointments
and treatments, within the local region, Canberra and Bega. Holders of a Veteran
Affairs gold card have access to taxis, with the fare paid in full by the Department of
Veterans’ Affairs.
69
3.2.4 Community services
Like most rural regions, Snowy Monaro is known for its many volunteer and
community organisations. The hospitals are supported through community auxiliaries
and volunteers. One organisation worthy of note, mainly due to its unique contribution
to the cancer community, is the Monaro Committee for Cancer Research Inc. (MCCR)4.
This organisation was started by a group of local women, after the death of a young
local mother in 1996. She had battled cancer for five years. The original aim was to
hold major fundraisers in the Monaro district with funds distributed to cancer research
and support for local cancer care facilities. Locally these funds have been used to
purchase equipment and provide financial assistance for cancer patients either
undergoing treatment or receiving palliative care. MCCR was a significant lobby force
behind the opening of an outpatient oncology unit in Cooma, and used their funds to
equip and furnish the unit. This has reduced the travelling burden for some oncology
patients, as they can alternate their treatments between Cooma and Canberra.
3.2.5 Snowy Monaro region summary
I hope by now the reader has a clear impression of the rural setting for this original
research. In some ways, it is a ‘typical’ Australian rural region, full of resilient, hardy
characters; and yet in many ways it is unique. It is a diverse region. Some areas, such as
Cooma, are well resourced and in relative proximity to a tertiary hospital. Yet other
locations, such as Delegate and Ingebyra on the Snowy River, are far from resources.
Some places are accessible in summer, but isolated in winter. Some places are easily
accessible by car, and yet for the weak and frail a 30-minute car journey is agonising.
3.3 The target population
The Snowy Monaro residents are a heterogeneous group. Some participants are sixth
and seventh generation Monaro farmers; others immigrated to Cooma during the
‘Snowy’ days; some followed their hearts during courting days; some came for work;
while others are new comers - they just decided, it would be a nice place to live (Figure
4 Monaro Committee for Cancer Research. Not for profit community fundraising organisation. Funds
cancer research, financial aid for local families coping with cancer, local health facilities e.g. equipment
provided to the oncology unit. <http://mccr.org.au>
70
3.9). The interviewed participants will be introduced in the following chapters. In the
remainder of this chapter I will discuss the challenges in recruiting palliative care
patients to research.
Figure 3.9: Great Diving Range, view from outskirts of Cooma
3.4 Challenges in recruiting palliative care patients to research
Challenges in recruiting palliative care patients to research have been well documented
in the literature (Agar et al., 2013; Rainsford et al., 2014; Wohleber et al., 2012).
Traditional views in palliative care have considered it unethical (De Raeve, 1994) to
recruit terminally ill patients to research, in the belief they are too vulnerable, too
unwell, too exhausted, or lack sufficient cognitive capacity to provide informed consent.
As dying patients have little to gain personally from participating, health care
professionals often become protective, acting as ‘gatekeepers’ to exclude these patients
in the belief they would find any request to participate as intrusive (Bullen et al., 2013;
Ewing et al., 2004). Members of human research ethics committees (HREC) also hold
reservations about providing approval to recruit such participants, and place strict
guidelines on how to proceed. However, recent studies argue this is not necessarily the
case (Ross & Cornbleet, 2003; White & Hardy, 2010). In response to the suggestion that
a recorded interview be stopped, due to the personal nature of the conversation, one
71
patient interviewed by Kleinman (1988, p. 148) replied, “No don’t. This helps me feel I
will leave something behind.” Gysels et al. (2012, p. 1) concluded in their literature
review that ‘ethical concerns regarding patient participation in end-of-life care research
are often unjustified’ and that providing studies are carefully designed and executed,
most palliative care patients and their FCGs find participation a ‘positive, therapeutic
experience’. This was certainly my experience during the interviews I conducted. One
bereaved FCG appreciated being involved. His wife had complex pain issues, and when
her cancer pain was finally brought under control by the specialist palliative care team
in Canberra, with a combination of second and third line medications, the local GPs
would not listen to her plea not to change the medications. No one had listened to his
wife. Theirs was a story “that needs to be told” (Andrew, husband of Elaine), and now
finally some months after her death he had an opportunity for “someone to hear [their]
story.” (Andrew).
However, concerns about ethics (Casarett, 2005) are not the only barrier to recruiting
palliative care patients. There are logistical issues that impede recruitment. The disease
trajectory of palliative care patients means they frequently present with complex
symptoms, including extreme mental and physical fatigue, and poor performance status.
Of those who do consent to participate in research, patient attrition becomes a hurdle, as
patients often die or become too unwell to continue in studies.
Recruiting FCGs can also be challenging. They are often too busy, stressed or
overwhelmed with the burden of care (Brazil et al., 2013). The research proposal,
discussed in the following chapter, had provision to conduct longitudinal interviews.
While this was possible with a couple of patients and their FCGs, the anticipated
number was not achieved. The daughter of one patient (Bruce) consented to be
interviewed but was not present at the initial interview with her parents. Her elderly
mother became distressed when Bruce, no longer unable to remain home for EoLC, was
admitted to a RACF. She lived on a farm and moved into town with her daughter, to
enable daily visits to her husband. The daughter became over-burdened with concern
for both parents, transporting her mother to and from home and the RACF, and running
a small business. On many occasions she stated to me, as her father’s palliative
medicine specialist, she was barely coping. It was inappropriate to ask her for an
interview.
72
One of the inclusion criteria for this study was that potential participants had some
knowledge of their prognosis. This also impeded recruitment as some patients who were
identified as suitable either did not know or did want to believe their life expectancy
was short. Some were so focused on controlling the disease they did not want to talk
about dying.
The challenges in recruiting palliative care patients to research are amplified in rural
settings. Urban studies have a larger pool of potential participants from which to recruit.
If a potential rural participant declines to participate, especially if they reside in the
more remote area, there is often a significant period to wait until another suitable
participant is identified. Despite the barriers, there is emerging evidence to suggest
conducting research in palliative care is feasible (Gibbons et al., 2013) especially if
strategies are implemented to ameliorate these problems during the development phase
of a research project (Bullen et al., 2013; Fischer et al., 2011).
3.5 Chapter summary
This chapter has described the setting for this research study, and outlined the
challenges faced in recruiting palliative care patients to research. The setting is a diverse
rural area. The following chapter outlines the methodology and methods chosen to
conduct this rural research study, and discusses the strategies used to overcome the
challenges faced in recruiting terminally ill patients. The reader will also be introduced
to the rural participants who are well placed to describe their experiences and
perspectives on rural end-of-life care, the ‘good death’ and place of death.
Figure 3.10: Monaro Dorper lambs
73
CHAPTER FOUR:
RESEARCH DESIGN AND METHODOLOGY
Introduction
This chapter introduces the research questions to be answered, and the methodology and
methods that underpin the research. The overall research plan explored the concept of
the good death, and the influence of place of death (PoD) and rural residency on the
‘good death’, from the experiences and perspectives of rural patients and their family
caregivers (FCGs). This was achieved by using two types of research (i) scoping and
systematic literature reviews, and (ii) mixed methods original research.
The overall project had three stages: preparation and ethics approval; data collection;
and analysis and write-up. The research objectives and questions guiding this thesis are
stated in Section 4.1. One element of the preparation stage was my literature review, as
discussed in Chapter 2. The preliminary literature review was enhanced by a more
systematic review of the literature, resulting in the completion and publication of one
scoping review (The rural good death: a scoping review) and two systematic literature
reviews (Place of death in rural palliative care: a systematic review; Rural end-of-life
care from the experiences and perspectives of patients and family caregivers: a
systematic review). Section 4.2 discusses the methods that guided these literature
reviews. In Section 4.3 I present my arguments for the theoretical framework and
chosen methodology underpinning the original research: mixed methods incorporating
ethnography and quantitative analysis. Section 4.4 outlines the qualitative methods,
including recruitment and consenting of participants, data collection (interviews), and
data analysis. Section 4.5 outlines the quantitative survey data collection and analysis.
In Section 4.6 I reflect on my role as a researcher, and the influence of my background
and perceptions on the data collection and analysis. A brief discussion on the limitations
of this study follows in Section 4.7, before concluding with my arguments that validate
the rigor of this research (Section 4.8).
74
4.1 Research objectives and questions
The primary objective of this research study was to explore the concept of the good
death from the experiences and perspectives of rural patients with a life limiting illness,
and their FCGs, within the rural Snowy Monaro region of NSW, Australia. The
secondary objectives were to determine the influence of PoD and rural residency on the
concept of the good death, and to explore the meaning behind the PoD preferences and
decision-making process. The hope was, that by identifying the essential features of the
‘good death’, and the meaning of PoD, in this rural setting, more informed decisions
and actions can be considered for inclusion in the palliative EoLC offered in this region.
The literature frequently reports the importance of placing qualitative data into context.
The perspectives obtained through the interviews were therefore contextualised by
describing the actual PoD of Snowy Monaro residents, and by exploring the factors
associated with PoD.
To address the objectives, six questions were explored through this research study:
1. What is a ‘good rural death’?
2. What is the influence of PoD on the ‘good rural death’?
3. What is the influence of rural residency on the ‘good death’?
4. What is the meaning of ‘place’ when deciding preferred PoD?
5. Where do Snowy Monaro residents die?
6. What factors influence the PoD? For example, are residents living on farms
less likely to die at home than residents living in town?
4.2 Systematic and scoping reviews
The starting point for this research was a general review of the literature, as reported in
Chapter Two, to identify gaps within the current literature that would help determine the
objectives of this study. While some might argue that systematic literature reviews are
not ‘research’, I have taken the stance that systematic reviews are in fact research, as
they answer a research question, and follow strict methodological guidelines for data
75
collection and reporting. At the 15th World Congress of the European Association for
Palliative Care (Madrid, May 2017), the Editor in Chief of Palliative Care, Professor
Catherine Walshe, on presenting the ‘2016 paper of the year award’ to the authors of a
systematic review, stated “reviews are research too”.
As previously noted, my initial literature review was sufficient to inform the
development of my research protocol and ethics application; however, due to the dearth
of identified rural studies, a more systematic approach was later undertaken, resulting in
one scoping and two systematic literature reviews. All three reviews were substantially
my own work; however, all my supervisors contributed to the publications, in varying
degrees, as stated on page viii of this thesis. The three reviews are now published and a
copy can be found in Appendix 6. I performed an update of the reviews (to 28th
February 2017) without any external assistance. The database updates identified all
three of my published reviews; however, my reviews were not included in the studies
added to the final literature review.
4.2.1 The rural ‘good death’: a scoping review
My initial intention had been to undertake a systematic review exploring the concept of
the rural good death, from the perspectives of rural residents, in order to answer the
research question “What is a ‘good rural death’?” However, due to the very small
number of relevant articles identified, a scoping review approach was chosen.
Currently, there is no standardised definition or methodology for scoping reviews
(Peters et al., 2015) so the definition commonly applied is that first described by Mays
et al. and cited and used by Arksey and O’Malley (2005, p. 5). They describe a scoping
review as one that aims:
To map rapidly the key concepts underpinning a research area and the main sources and
types of evidence available, and can be undertaken as stand-alone projects in their own
right, especially where an area is complex or has not been reviewed comprehensively
before.
The protocol used in my review was based on the methodological framework first
described by Arksey and O’Malley (2005), enhanced by Levac et al. (2010), Daudt et al.
76
(2013) and Colquohoun et al. (2014), and later refined by the Joanna Briggs Institute
(JBI, 2015). The original (Arksey & O’Malley, 2005) and modified (Levac et al., 2010)
frameworks consist of six stages (JBI, 2015):
1. Identifying the research question by clarifying and linking the purpose and
research question.
2. Identifying relevant studies by balancing feasibility with breadth and
comprehensiveness of the scoping process.
3. Study selection using an iterative team approach to selecting studies and
extracting data.
4. Charting the data by incorporating a numerical summary and qualitative
thematic analysis.
5. Collating, summarising and reporting the results, including identifying the
implications for policy, practice and research.
6. An optional consultation exercise between stakeholders.
In recent years, scoping reviews have become increasingly popular. To ensure scoping
reviews are validate and reliable, Tricco et al. (2016) have identified the need for
standardised reporting guidelines. However, at the time of my scoping review no
guidelines existed. Therefore, the reporting of my scoping review was based on the
2016 scoping review published by Tricco et al., the team developing the standardised
reporting guidelines, the Preferred Reporting Items for Systematic Reviews and Meta-
Analysis-Scoping Review (PRISMA-ScR) guidelines (Equator Network, 2016). While
quality assessment is not a requirement of the JBI guidelines, it was included in my
published review (Levac et al., 2010; Daudt et al., 2013); however, the optional
consultation stage was omitted.
4.2.1.1 Eligibility criteria
Scoping reviews have a broad approach and include any existing literature, including
both published (primary research studies, observational studies, reviews) and relevant
unpublished (grey literature) articles, regardless of study design, discipline or quality.
To answer my scoping review research question, I developed eligibility criteria, using
the Participants, Concept and Context (PCC) acronym outlined in the JBI guidelines
(2015):
77
Participants: Rural residents, including rural patients with a life-limiting illness,
rural FCGs, rural HCPs and the wider rural community, as these are the most
appropriate to provide the rural perspective. No age filter was applied.
Concept: The principal concept under review was the ‘good death’ in a rural
setting, as described through the personal experiences or perspectives of rural
residents; collected by interviews, surveys or extensive field work observations.
The term ‘good death’ was either used explicitly or implied.
Context: Rural or remote; all countries and territories were considered; no
standardised definition of rurality was used. Articles that included urban and
rural data were considered providing the rural data were clearly identifiable.
4.2.1.2 Information sources and search strategy
Five electronic databases (PubMed, CINAHL, Scopus, PsycINFO, and Web of Science)
were searched from 2nd January through 14th February 2016. A literature review update
was undertaken in early March 2017, by accessing the same data bases and searching
for articles published from 15th February 2016 to 28th February 2017. An iterative
process was used with “peaceful death” omitted from the original search. The following
keywords and Medical Subject Heading (MeSH) terms were used in the final search:
(“good death” OR “managed death” OR “good enough death” OR “tamed death” OR
“dying well” OR “peaceful death”) AND (Rural OR Remote). All study designs were
included; no date filter was applied; only English language papers were included.
The initial search identified 377 articles. These were downloaded to ENDNOTE X7,
merged and duplicates deleted (338 articles). The reference lists of all retained articles
were scanned for additional studies. Recent issues (July 2014 - Jan 2016) of eight
relevant journals (Palliative Medicine, Journal of Palliative Medicine, Palliative and
Supportive Care, Australian Journal of Rural Health, Journal of Rural Health, Social
Science and Medicine, Health and Place, and Death and Dying) were hand searched.
Cochrane Library, CareSearch database and OpenGrey repository were searched for
grey literature. Authors of three studies reporting mixed geographical data were
contacted; however, rural data were not specified, and so these three articles were
excluded from the review. The literature search update followed the same format and
identified 19 papers, including my own published paper. On review of titles and
78
abstracts no suitable papers were found to add to this scoping review. The remainder of
the selection process is illustrated in the flow chart (Figure 4.1).
Figure 4.1. Flow diagram of scoping review selection process including reasons for exclusion.
(Source: modified flow chart as described by Moher et al., 2009)
4.2.1.3 Assessment of quality
To ensure valuable insights reported in lower quality studies are not excluded, the
current scoping review methodological guidelines do not require a formal quality
assessment of eligible articles. However, a quality assessment was conducted in this
review to assist in validating the quality of the literature informing my report. After an
initial assessment was made, my four supervisors independently assessed each paper.
Differences were discussed by email and resolved by consensus. All eligible articles
Phase 1 screening♦: Titles reviewed
independently by SR and one supervisor
Articles after searches merged, duplicates
removed (n =338)
Irrelevant (n=290)
Phase 2 screening♦: Abstract and/or full-
text articles assessed independently for
eligibility by SR and supervisors (n=48)
Articles identified through database
searching (n =377)
Articles excluded (n =35)
Not good death (16)
Not rural (2)
Rural data not identified (4)
Not relevant (4)
Duplicate [poster] (2)
Workshop abstract (1)
No personal perspectives (6)
Eligible studies (n =13)
Studies added through hand search of
references and journals (n=7); Grey
literature search (n=0)
Articles included in scoping review
(n=19) reporting 17 separate studies.
(15 qualitative, 1 quantitative,
1 mixed methods.)
Literature search update; 19 papers
identified; nil suitable to add to the review.
Systematic
review (n=1) ♦ Differences were discussed via email
and resolved by consensus
79
were retained in this scoping review, regardless of their quality. Some high-quality
studies received a lower score, as the assessment was based on aspects relevant to the
rural ‘good death’, and not of the study per se. All studies identified for the scoping
review were rated to be of low, medium or high quality based on a simple scoring
system described by Hawker et al. (2002) and modified by Gomes et al. (2013). Two
additional items were added to account for the rural ‘good death’ focus: (1) clarity of
rural definition and (2) validity of informant (prospective=2, retrospective =1, well
community=0). The scoring card can be found in Appendix 2.
4.2.1.4 Synthesis
The synthesis included both a quantitative analysis of the actual scoping review and a
qualitative analysis of the content of the included articles (Appendix 3). A thematic
analysis of the content was conducted by downloading the eligible articles into NVivo-
10, and coding for major themes. Findings were reported narratively. Due to
heterogeneity within a small number of studies, a meta-analysis and analysis according
to informants were not possible. Some informant groups had only one study identified.
The qualitative results of the scoping review were described in Chapter Two (Section
2.4). The published paper can be found in Appendix 6.
4.2.2 Systematic reviews
Two systematic literature reviews were conducted identifying both quantitative and
qualitative studies, and were undertaken utilising the Preferred Reporting Items for
Systematic Reviews and Meta-Analysis (PRISMA) guidelines (Shamseer et al., 2015).
4.2.2.1 Place of death in rural palliative care: a systematic review
The first systematic literature review, “Place of death in rural palliative care: a
systematic review” was conducted between September and December 2014, by
searching four electronic data bases (PubMed, PsycINFO, Scopus, and CINAHL) using
the following keywords and Medical Subject Heading (MeSH) terms: (choice OR
prefer* OR decision) AND (palliative OR end-of-life OR terminal) AND (care OR
death) AND (place OR location OR site) AND (Rural). A date filter was not applied,
and all published English language, peer reviewed research articles were identified.
Having found 82 manuscripts, the search was widened by limiting the keywords to
(“place of death” OR “place of care”) AND (Rural), on the presumption rural data may
80
be embedded in population studies. The two searches were downloaded to ENDNOTE
X7, merged and duplicates deleted. The reference lists of all included studies were
scanned for additional articles. The literature search was updated in early March 2017,
using the same databases and keywords, and date limitation of January 2015 through
February 2017. Issues (June 2013 - February 2017) of six relevant journals (Palliative
Medicine, Journal of Palliative Medicine, Journal of Pain and Symptom Management,
Palliative and Supportive Care, Australian Journal of Rural Health, and Journal of Rural
Health) were hand searched by myself. Relevant articles, found incidentally or
identified by PubMed monthly updates (January 2015 – February 2017), were also
included. One corresponding author (Thomas et al., 2003) provided a copy of their
completed report containing specific rural data. One hundred and thirteen articles were
identified in the original search, with 40 being eligible for inclusion. An additional 38
papers were identified in the database search update. After removing duplicates and
those irrelevant by title and abstract, four additional papers were included (Figure 4.2).
Selection criteria
Each identified study was assessed against predetermined inclusion criteria, and were
included if they fulfilled the following conditions:
Preferred and/or actual place of death/care reported as either part of the
research aim or findings.
Participants included rural, regional or remote residents (no standardised
definition of rurality was used).
Rural data were clearly identifiable.
Cause of death included a palliative condition (malignant and non-
malignant), or the survey included preferences indicating the participants
had current or hypothetical life-limiting illness.
The article indicated a research study had been conducted with method
and results described.
While this systematic review excluded paediatric palliative care the age
filter (≥19years) was not applied in the initial search as different cut off
ages were used, ranging from 15 to 19 years.
Both qualitative and quantitative studies were included.
81
Studies were described in terms of country, date of publication, populations (including
palliative condition), design, rural description and assessment outcome. Two main
categories were identified, actual place of death (APoD) and preferred place of care
and/or death (PPoC/PPoD), with results summarised in two separate tables.
A quality assessment of each paper was conducted. One of my supervisors
independently assessed a random selection of six studies (17%). Differences were
discussed by email and resolved by consensus. Different scales were used to assess the
quality of the quantitative and qualitative studies. Mixed studies were assessed using
both scales. The quality assessment tool outlined previously in the scoping review, was
used for the qualitative studies (Appendix 2). The scale suggested by Khan (2001, as
cited in Gomes & Higginson, 2006), and modified by Gomes et al. (2013) was used for
the quantitative studies.
Quantitative data were manipulated into a common format and where possible,
algebraic back calculations were performed to determine the percentage of actual or
preferred rural home deaths. Where the adjusted odds ratio (aOR) was reported, the rate
of a rural home death was compared to urban (aOR of urban home death=1). Thematical
analysis of the qualitative data regarding PoC/PoD are reported descriptively. Due to
heterogeneity, a meta-analysis was not possible. Findings are reported in Chapter 2
(Table 2.2 and Table 2.3 summarise the included studies). The full data extraction tables
can be found in the published paper in Appendix 6.
82
Figure 4.2: Flow diagram of selection process for 'Place of death in rural palliative care: A
systematic review.'
PoD: place of death; PoC: place of care; APoD: actual place of death; PPoC/PPoD: preferred place of care/death.
♦ Differences were discussed via email and resolved by consensus
4.2.2.2 Rural end-of-life care from the experiences and perspectives of patients and
family caregivers: a systematic review
The second systematic literature review, “Rural end-of-life care from the experiences
and perspectives of patients and family caregivers” was conducted in January 2016,
again by searching four electronic databases (PubMed, CINAHL, Scopus and Web of
Science) using the following keywords and Medical Subject Heading (MeSH) terms:
(“palliative care” OR hospice OR “terminal care” OR end-of-life OR “end-of-life care”)
AND (Rural OR Remote) AND (patient OR family OR carer OR caregiver) AND
(perspective OR experience OR opinion OR view). An initial date filter of January 2006
n=4
Articles identified through database
searching (comprehensive)
(n = 82)
Articles identified through ‘PoD or
PoC AND RURAL’ database
searching (n = 172)
Articles after searches merged, duplicates removed (n = 96)
Articles excluded (n = 46)
Paediatric (8); maternal
mortality (4); Education (3);
Trauma/suicide (6); advance
care planning (2); service
provision (6); irrelevant (17)
Titles reviewed by SR and
one supervisor♦
Abstract and/or full-text
articles assessed for
eligibility by SR and one
supervisor (n = 50)♦
Articles excluded (n =27)
[No separate rural data (7);
not rural (1); not PoC or
PoD (16); no separate
‘palliative’ rural data (2);
Non-English (1)]
Eligible studies (n = 23) Studies added through
hand search of references,
journals and PubMed monthly update, author
contact (n=17) Articles included in systematic
review (n=44) reporting 39
separate studies
Studies identified in search
update:
Database searches (n=37)
Journal review (n=1)
(34 not suitable)
APoD studies n=26
PPoC/PPoD
studies
n=11
APoD/PPoD
studies
n=2
83
through January 2016 was applied to include an overlap with the last published
systematic review (C Robinson et al. 2009). This allowed for any missing submitted,
but unpublished studies to be included. One article (Payne et al. 2007) was identified
and included.
The reference lists of all included studies were scanned for additional articles. A hand
search of recent issues (July 2014 - January 2016) of six relevant journals (Palliative
Medicine, Journal of Palliative Medicine, Journal of Pain and Symptom Management,
Palliative and Supportive Care, Australian Journal of Rural Health, and Journal of Rural
Health) was also undertaken. An updated search of the previous databases and journals
was undertaken, in early March 2017, using a date filter of 1 February 2016 to 28
February 2017. Sixty-six additional potential papers were identified in the databases,
including two of my published papers. No relevant rural studies were identified in the
updated hand search of journals (Figure 4.3).
Studies were included if they were consistent with the following inclusion criteria:
Rural or remote residents (no standardised definition of rurality was
used).
Participants were receiving palliative or EoLC for malignant or non-
malignant illnesses.
Data pertained to patients’ or FCGs’ experiences or perspectives on
EoLC, collected from interviews or surveys.
Both qualitative and quantitative studies were included.
No age filter was applied.
Only published international English language, peer reviewed research
articles were considered. Articles that included urban and rural data were
included, providing the rural data were clearly identifiable.
Rejected studies were either clearly irrelevant or those that addressed the topic in
general but failed in one or more of the inclusion criteria. After an initial quality
assessment was made, verification was shared independently by the five co-authors. The
previously outlined scoring card was used. Differences were discussed by email and
resolved by consensus. While low quality studies were excluded from the published
systematic review, all studies, regardless of quality, were retained for my thesis
literature review (Table 2.4, Chapter 2).
84
Study findings were coded into four categories: (1) patient perspectives; (2) FCGs
perspectives; (3) facilitators and barriers to receiving rural palliative EoLC; (4)
influence of rural place and culture on EoLC. Each category was analysed thematically
and reported descriptively.
Figure 4.3: Search flow chart ‘Rural end-of-life care from the experiences and perspectives of
patients’ and family caregivers’: A systematic review’.
Articles identified through database
searching (n =442)
Articles after duplicates removed
(n =212)
Titles screened independently by SR and
one supervisor (n=220)♦
Irrelevant (n=115)
Abstract and/or full-text articles assessed
independently for eligibility by SR and
supervisors (n=105)♦
Additional articles identified through hand
searches (n =8)
Articles excluded (n =75)
Not patient/FCG (28)
Patient/FCG not separated (4)
Not rural (7)
Rural data not separate (9)
Service evaluation (10)
Not EoLC (6)
Not a study (5)
Other (6)
Eligible studies (n =30)
(n=2)
Articles included in systematic review
(n=32) reporting 27 separate studies
Database search update studies n=66 of
which 64 were not relevant
Systematic
review
(n =1)
Medium
quality studies
(n=12)
High quality
studies
(n =11)
Source: modified flow chart as described by Moher et al. (2009) Abbreviations: FCG- family caregiver;
EoLC- end-of-life care ♦ Differences were discussed via email and resolved by consensus
Low quality
studies
(n=3)
85
4.3 Theoretical framework and methodology of the mixed methods
research study
While previous studies have utilised written questionnaires and surveys (Patrick et al.,
2001), such as the ‘Quality of death and dying questionnaire’ (Curtis et al., 2000), in-
depth qualitative interviews are more appropriate to investigate the participants’
perspectives and viewpoints on the ‘good death’ as:
One of the great advantages of qualitative methods is that they enhance the capacity not
only to describe events but to understand how and why the same events are often
interpreted in a different, sometimes even conflicting manner, by different stakeholders
(Sofaer, 1999, p. 1106).
The secondary aims of this research were to determine the influence of PoD, and rural
residency on the concept of the good death, and to explore the meaning behind the PoD
preferences and decision-making process. Again, this information is best captured in
open-ended interviews. To complement the fieldwork, and to place the interview data
into context, I also completed a descriptive study of current practice, addressing the
questions ‘Where do Snowy Monaro residents die?’ and ‘What factors influence place
of death in this rural setting?’ My research therefore, used a mixed methods approach,
combining ethnography methods (detailed in Section 4.4), to gain a deeper
understanding of, and the reasoning behind the actual or preferred PoD, with descriptive
analyses from the survey data (detailed in Section 4.5).
The theoretical perspective, or philosophical stance, informing the methodology and
epistemology (the theory of knowledge embedded in the theoretical perspective, and
thereby in the methodology) is symbolic interactionism, developed by Herbert Mead
(1863 -1931), in which people create meaning based on their interactions with others.
The meaning of events and behaviours arises from an individual’s interpretation.
Behaviour, therefore, is not necessarily based on objective truth, but on what a person
believes to be true. Interpretations of the same event can vary from person to person
(Crotty, 1998). Symbolic interactionism, in turn, reflects the fundamental
epistemological stance adopted by this research, social constructionism:
86
Truth or meaning comes into existence in and out of our engagement with the realities
in our world…. meaning is not discovered, but constructed…different people may
construct meaning in different ways (Crotty, 1998, pp. 8-9).
Constructionism is based on the view that reality is based on knowledge. Knowledge is
formulated through social context, and it is this context that dictates the way people
view or construct ‘their world’. Interactions between people and ‘their world’ constructs
meaningful reality. Therefore, there is no single or true interpretation of events, only
meaningful or useful interpretations. Social constructionism infers that ‘without culture
we could not function…. We depend on culture to direct our behaviour and organise our
experience’ (Crotty, 1998, p. 53). The theoretical framework is illustrated in Figure 4.4.
Figure 4.4: Theoretical framework and methods underpinning the complete research study.
4.3.1 Ethnography
Qualitative research, utilising interviews and observation, is interpretative and studies
‘things in their natural settings, attempting to make sense of, or interpret, phenomena in
terms of the meaning people bring to them’ (Denzin and Lincoln, 2005, as cited in
Epistemology
Theoretical
Framework
Methodology
Social Constructionism
Symbolic interactionism
Methods
1. Literature review
2. In-depth interviews
3. Survey data descriptive analysis
Mixed methods
Systematic and scoping literature reviews
Ethnography
Quantitative survey
87
Grbich, 2015, p.40). There are perceived differences between rural and urban cultures,
and the influence of rural culture on the ‘good death’ and place of care/place of death is
best explored through an ethnographic study. Ethnography specifically ‘deepens the
understandings of cultures and culturally based concepts’ (Wilson et al., 2009a, p. 21).
In this study, ethnography provided deep insight into the participants’ perspectives on
the rural ‘good death’, the meaning behind PoD preferences and decision making, and
the influence of rural culture and values on the ‘good death’ and PoD within the Snowy
Monaro region.
Ethnography, a research method developed by anthropologists to describe and
understand foreign cultures, is used to understand groups of people living within society
(e.g. the marginalised or invisible). As summarised by Patton (2002, p. 81), ‘The
primary method of ethnographers is participant observation… [along with] extensive
fieldwork in which the investigator is immersed in the culture under study.’ Variations
in cultures are recognised by gaining and describing ‘the understandings and meanings
constructed by people as they undertake daily activities’ (Grbich, 1999, p. 159).
Becker et al. (2004, p. 269) suggest that traditional ethnography defines ‘social
problems that have either not been recognised or have fallen off the radar
screen…because the presence of this problem contradicts cultural norm.’ In this study
‘cultural norm’ is the culture of palliative care, as studied from an urban perspective,
where urban residents requiring EoLC usually have access to a specialist
multidisciplinary palliative care team. Becker et al. (2004, p. 271) also argue that:
Ethnography has the capacity to develop different ways of thinking about a social
universe that is often taken for granted, to explore and make real the experience of
people who have gone through hardship, through loss…and possibly even a triumph
against all expectations.
Within qualitative research there are two main schools of thought regarding the
positioning of the researcher. Grbich (1999, p. 5) suggests some would say ‘the
researcher can participate in, and document, the ‘outer’ world with minimal intrusion’
while others believe ‘the perceptions of the researcher and other participants in the
‘socially constructed’ world are intricately interwoven’. Ethnographic researchers are
88
conscious of the potential for bias in all research, as every aspect, starting from the
research question, design, methodology, selection, analysis, conclusions and responses
to ‘ethically important moments’ (Guillemin, 2004), are influenced by our values and
beliefs, background and motives, and are open to subjective interpretation (Ulin et al.,
2005). Essential components of ethnography are transparency and reflexivity (Rice et
al., 1999). Reflexivity is a process of critical reflection by the researcher. It is important
that the researcher is able to stand back, at every stage of the research, and take a critical
look at their own role in the whole research process.
Researcher influence is not just limited to my personal world view, but my very
presence within the study, and my personal framing. ‘Rigorous qualitative research is
honest about the role of the researcher in the project’ (Rice et al., 1999, p. 41); however,
despite all attempts, no researcher can be totally unbiased. Steps (described later) were
taken in my research, starting at recruitment through to the final writing, to minimise
my potential bias in collecting, analysing and reporting the research results.
4.4 Ethnographic component
A qualitative approach, utilising open-ended interviews as the prime method of data
collection, was chosen to ensure there was sufficient depth and detail to fully
understand each participants’ personal experience (Patton, 2002, pp. 16-17), and to
enlighten the interpretations of the quantitative data. While there is limited data on
EoLC from the experiences and perspectives of patients and their FCGs, I felt it was
important not to exclude the local rural health professionals (palliative care nurse
[PCN], community nurse [CN], residential aged care facility director of nursing [RACF
DON], and general practitioners [GPs]), as a cross section of viewpoints would
facilitate triangulation and thus provide a more accurate picture of the realities of the
rural ‘good death’.
4.4.1 Recruitment
To ensure a cross-section of experiences and perspectives, three groups of participants
were recruited for the ethnographic study: patient and FCG dyads; local health care
professionals (HCPs); and bereaved FCGs, whose family member had died up to twelve
months previously (Figure 4.5).
89
Figure 4.5: Recruitment flow chart for interviews
Recruitment for interviews was conducted through direct invitation. Twenty seven
potential patient participants, who fulfilled the inclusion criteria (Table 4.1) and
triggered a “no” response to the ‘surprise question’, “would you (the referrer) be
surprised if this patient were to die within the next 6 months?” (Gold Standard
Framework, 2011), were identified by the palliative care nurse (19), community nurse
(1), one RACF DON (2), two GPs (3) and myself (2). Except for the purposely selected
bereaved family members, who received their invitation by mail, most FCG participants
were present at the time of patient invitation. The three absent FCGs were identified by
the patient, and with patient consent, were contacted at a later time. Over time,
recruitment became more purposeful, as I tried to ensure a cross-section of participants
representing all three shires, different degrees of rurality, both cancer and non-cancer
diagnoses, and those receiving care at home, hospital and in a RACF.
Patients identified, n=27
Patients ineligible, n=3
Died or deteriorated prior
to consent, n=2
Patients refused, n=8
Patients consent (verbal), n=14
Died /deteriorated prior to
interview, n=3
Patients interviewed
(written consent),
n=11
Family care giver, n=15
Separate bereavement
family care giver, n=3 Health care professional, n=6
PCN, n=1
CN, n=2
DON, n=1
GP, n=2
Total participants, n=35
Abbreviations: PCN- palliative care nurse; CN- community nurse; DON- residential aged care facility
director of nursing; GP- general practitioner.
*For 1 consenting patient the FCGs were interviewed first, but the patient became too unwell before they could be interviewed
90
Table 4.1 Inclusion criteria for interviews
Patient Family caregiver Bereaved family caregiver
- Life limiting illness
- Life expectancy ≤6 months
- Awareness of prognosis
- ≥18 years
- Snowy Monaro resident
- English speaking
- Capacity to provide consent
- Family member or person
identified by patient as primary
informal caregiver
- ≥18 years
- English speaking
- Capacity to provide consent
- Next of kin identified in
obituary notice
- Deceased relative was a
Snowy Monaro resident
- Relative died of a life-limiting
illness
- ≥18 years
- English speaking
- Capacity to provide consent
Before commencing any interviews, the objectives and processes of the research were
discussed with the PCN, CN, DON and a principal GP in each of the general practices. I
personally made a presentation during a lunchtime meeting to the two GP practices in
Cooma. The nurses and GPs were phoned, and invited to be interviewed, after the
patient had died. Strategies for maintaining confidentiality were discussed (real names
and ages have not be used in any report). Recruitment strategies ensured voluntary
participation. Conflict of interest was addressed by advising each GP when one of their
patients had consented to participate. This was not just a courtesy, but provided
opportunity to assure the GP I would not be taking over the care of their patient, unless I
received a formal referral at a later date.
In accordance with the ANU and GWNSW LHD ethics requirements, on the two
occasions where I directly approached one of my current patients to invite them to
participate, a third party (practice nurse) conducted the consenting process to ensure
there was no coercion, on my part, to participate. At the time of recruitment, one of my
current patients identified by another GP, declined to participate.
Once identified by the recruiter, the patient gave verbal consent to be contacted directly
by phone. I responded to all identified potential participants who had given verbal
consent, with the exception of two who died unexpectedly, prior to making contact. The
reasons for not participating included not wanting to talk about dying, and being too
busy travelling for treatment. One Asian FCG of an interested Anglo-Saxon patient,
identified by the palliative care nurse, hid the paperwork from the patient as she felt it
was culturally inappropriate to talk about dying.
91
4.4.2. Participant consent
Before commencing interviews, participants were provided with a detailed participant
information sheet (a copy can be found in Appendix 4), outlining the objectives of the
study, the interview process, the topics to be covered, and the way their information
would be managed and stored. Participants were assured they were free to withdraw at
any time, up to the time of thesis submission or journal publication. Participants
completed a signed consent form before commencing the interview. An iterative process
of consent was undertaken. At the start of each interview I gained verbal consent to
proceed, and at the end of each interview the participant was asked if there were any
details they wanted deleted. At the start of each interview, the patient and their FCG
were reminded that I was interviewing them as a researcher, and not as their treating
doctor. As such, the participants were not to assume I knew anything about their illness
and its management; that I would not be providing advice; and that I was not implying
there would be a change in their clinical management by asking certain questions. For
example “Have you considered moving to a nursing home?” did not imply that was my
intention as their treating doctor. Despite my best intentions to allay patients’ concerns,
one patient did become upset, as he thought I was trying to send him to a nursing home.
His GP spoke to him at length the following day, and assured him he was remaining in
the hospital for EoLC. In other interviews, a few patients and FCGs teared up when
telling their story. In these instances, I gave them time in quietness, sometimes pausing
the recording. I was careful to ensure they were ready, and agreeable, to continue the
interview. The information sheet provided the contact details of a counselling service
(Beyond Blue) if they felt they required additional support. To my knowledge this was
not required by any participant, and in fact a number of bereaved FCGs commented on
how therapeutic it had been to talk about the death of their relative.
Between three and six months after the patient had died, all FCGs who had previously
consented were contacted by either phone or mail, to arrange a follow-up bereavement
interview. In respect for the FCGs’ grieving, they were not contacted immediately after
the death. However, the literature reports that if too much time has elapsed, then issues
of memory may impact the validity of the recall (Addington-Hall & McPherson, 2001).
Despite all FCGs previously consenting, this aspect of the research proved to be
difficult. Of the 15 FCGs, only five agreed to be interviewed, with another who talked
openly about his experience, during my follow up phone call, but declined a formal
92
interview appointment. My failure to contact the remaining bereaved FCGs, (two had
moved away, two did not return my phone calls, two had returned to their interstate and
overseas homes, and the remainder were uncontactable) is consistent with previous
studies (Bakitas, 2008, p. 7). Two bereavement interviews were face to face, the
remaining three were conducted over the phone.
While patients and FCGs are the main group ‘missing’ in rural studies, I felt it
important to also gain the views of the nurses and GPs directly involved in the care of
the patient. The PCN was aware of who had consented to participate as she was the
primary recruiter. The relevant GPs had also been contacted, to advise them of their
patient’s participation. After the patient had died, I again contacted the nurse and GP,
and invited them to be interviewed, for them to express their perspective on the EoLC,
death and PoD. Two GPs agreed, however a third did not respond to my invitation. I
was the primary treating doctor for three patients. At least one nurse or DON agreed to
be interviewed for each patient.
The third group of participants was a purposely selected group of bereaved relatives,
chosen because their story was possibly outside the stories of the recruited patients.
During my patient interviews none had died in the out-of-town hospice, so my search of
the obituary notices selected three FCGs whose family member had died in the
Canberra hospice. I also selected three FCGs whose family member had lived on a farm
and died in the local hospital; one whose wife lived in a village and died in the Bombala
MPS; a younger man who was cared for and died of a palliative illness at home; and an
elderly man who was transferred from the tertiary hospital to die in the local rural
hospital. Of nine invitees, four responded, with one regretfully advising that they were
still grieving after 12 months, and unable to participate.
All consenting participants are residents of small rural communities, and it is possible,
despite the use of de-identified data and aliases, they might be recognised through their
role, stories and quotes. Permission to include quotes and stories was included in the
consent process.
93
4.4.3 Interview participants
Table 4.2 lists all the participants, and describes the demographic features of the
patients (interviewed and identified in obituary notices). Names and ages have been
changed to preserve participants’ anonymity, though ages are within five years of the
true age.
4.4.4 Data collection - the interviews
This research used open-ended interviews. I refrained from using a structured interview
format, as I wanted to explore deeply the views, perspectives and experiences of the
participants (Patton, 2002, p. 4), and wanted to be sufficiently flexible to allow the
participants to set the direction of the discussion. I did however, have a set of prompt
questions, adapted from those described by McCall and Rice (2005) (Appendix 5) to
assist in the interview process. On a few occasions, I referred to the prompts early,
especially if the interviewed stalled; however, it was more common for me to glance
over the list at the end of the interview, to ensure all major points had been addressed.
The interview topics were based around the patients’ illness, care, needs and support
(what was good, what could be done better), where they would prefer to be cared for in
the last weeks and days, and if they had a preference for PoD. As the interviewing stage
progressed, my approach to the interviews became more iterative and more focused, as I
would specifically ask questions surrounding an issue raised by a previous participant. I
wanted to determine if the issue raised was specific to one participant or more
generalised. As I became more confident, and where appropriate, I asked specific
questions relating to the ‘good death’. “You have no doubt heard of someone living a
good life, what do you think would make a good death?”, and in the bereavement
interviews “did….die well?” or “did … have a good death?” In some interviews I asked
more general questions around the topic, as I had a sense the concept may not have been
readily received. I also asked about the benefits and challenges of rural residency on
EoLC and PoD.
94
Table 4.2: List of participants and demographic features of the patients (interviewed and
identified in obituary notices).
Patient Sex Age
Marital Diagnosis
Residence
Place of
death
Interviewed
family care-
giver
Interviewed
health care
professional
#1
Gwen
F 81 Widowed CRC Town RACF-
respite
Daughter
(Louise);
Lived 500km
away
DON
#2
Cheryl
F 56 De facto GBM Town,
RACF
RACF
(LT)
Partner (Pat) NIL
#3
Marianne
F 62 Married Myeloma Farm Home Husband
(Peter)
GP , CN
#4
Kevin
M 75 Married Sarcoma Town Home Wife
(Deedee)
Daughter
(Susan)
PCN
#5
Ray
M 74 Married Lung ca Town CDH Wife (Clare) PCN
#6
Ryan
M 56 Married Lung ca Rural
residential
Home Wife (Carol) PCN
#7
Barbara
F 92 Widowed Frailty Town -
RACF
RACF
(LT)
Son (Rodney)
Lived 400km
away
NIL
#8
Bruce
M 85 Married CRC Farm RACF
EoLC
Wife (Diane) NIL
#9
Dorothy
**
F 84 Married Pulmonary
Fibrosis
Town-
RACF
RACF
(LT)
Daughters x2
(Rhonda,
Simone)
NIL
#10
Ivan
M 83 Widowed Leukaemia Farm Home Son (Paul);
daughter
(Amanda-
interstate)
GP
CN
#11
Stuart
M 55 Single Pancreatic
ca
Town BMPS Brother
(Fred); lived
500km away;
retires and
stayed with
Stuart for last
weeks)
PCN
#12
George
M 62 Divorced Lung ca Village BMPS Sister (Betty);
lived in
separate
dwelling on
property
NIL
#13
Leanne
*
F 55 Partner Breast ca Farm CDH Partner
(Daniel)
NIL
#14
Elaine*
F 60 Married Breast ca Town Hospice Husband
(Andrew)
NIL
#15
Gordon
*
M 72 Married;
wife in
RACF
Cerebral
lymphoma
Town CDH; transfer
from
TCH for
EoLC
Son
(Nathan)
NIL
* Patient not interviewed (identified in obituary notice); ** Patient not interviewed as too unwell;
Abbreviations: RACF=residential aged care facility; LT= long term (permanent resident); HCP= health care
professional; CRC=colorectal cancer; GBM= Glioblastoma multiform; DON= director of nursing; GP=general
practitioner; CN=community nurse; PCN= palliative care nurse; CDH= Cooma hospital; ca= cancer;
BMPS=Bombala multipurpose service; TCH=The Canberra Hospital; EoLC= end-of-life care.
95
Where possible, interviews were conducted at the place of residence (private home or
RACF); however, a few interviews were conducted in hospital. On one occasion, a
patient was admitted to hospital on the morning of my pre-arranged interview. Despite
my hesitancy to interview the patient and his wife, on what was a distressing day, both
were eager to keep to the interview schedule. I kept the interview with this patient as
short as possible (15minutes). Two bereavement interviews, identified through obituary
notices, were conducted at my place of work. I had not met the men before and felt it
more professional to conduct the interviews in a more ‘public’ space. With the
exception of three bereavement phone interviews, all interviews were conducted face to
face, as meaning is not just the words but also the non-verbal clues, the facial
expressions, gestures in response to the words spoken, details that can only be observed
in interviews. This meant for one outlier I travelled 270 km round trip from my home to
conduct the interview.
In total, 42 interviews, lasting a total of 22.8 hours, were conducted, between 20th April
2015 and 24th November 2016. The interviews lasted from six to 78 minutes, (average
33 mins). The time spent with patients was determined by their physical condition.
Time spent with FCGs was determined by their availability. For example, a follow up
interview with Gwen’s daughter lasted just eight minutes, as she was concerned about
leaving her mother on her own. The interviews were mostly individual, with two joint
interviews with both patient and FCG present. I had intended to conduct a follow up
interview with each patient; however, most patients became too unwell, or it was not
possible to find a mutually convenient time, especially those living further away from
Cooma. Two patients died while I was overseas on holidays.
At the start of each interview, I re-affirmed that the interviewee consented to participate,
and for the interview to be audio recorded. Participants were reminded they could stop
the interview at any time. I did not take any notes during the interview; however, wrote
an entry in my personal journal (my feelings, the environment, physical condition of the
patient, interruptions) once back in my car.
The interviews were conversational and open-ended, however, became more focused as
I wanted to explore further issues raised in previous interviews. At the completion of the
interviews I gave the participants’ opportunity to raise any issues they felt had not been
96
covered, and to delete anything they did not want recorded. All interviews were audio
recorded, and transcribed verbatim. After transcribing, participants were assigned an
alias. Recorded interviews were deleted from the recorder after transcribing.
Downloaded interviews, transcriptions and notes are kept on my password protected
laptop. A backup copy is kept on an external hard drive. As per the standard ANU
policy, these files will be kept for five years after publication.
4.4.5 Analysis of qualitative data
Transcribing was a time consuming process, made more difficult by my elementary
typing skills. Transcribing of a one hour interview took up to eight hours to complete;
however, it provided opportunity to become embedded in the data and reminded me of
the emphasis placed on words, and the emotions displayed during the interviews. It was
during the transcribing process that I began initial analysis. However, after transcribing
the first five interviews I employed a known and trusted local online transcribing
service. On receipt of the transcripts, interviews were listened to and errors corrected.
The transcripts were re-read, then down loaded to NViVo-10. Data analysis was
undertaken using thematic analysis, which involved the identification, analysis and
reporting of patterns, themes and interrelationships within the data (Patton, 2002, p. 41).
In the two occasions I was able to conduct follow-up interviews, issues clouded in
uncertainty were clarified.
After becoming intimately familiar with the data, a coding tree was developed. Often
qualitative research is conducted by a team, where members would independently code
the data, then compare and discuss similarities and differences, in order to gain a
consensus. However, as this research is a PhD thesis, the coding was substantially
carried out on my own, with general guidance provided by one supervisor with
extensive experience in ethnography. After determining the three main themes (place,
rural/community, and ‘good death’), coding was inductive (discovering patterns and
themes) rather than deductive (analysing data according to a predetermined framework)
(Patton, 2002, p. 453). Each major theme was classified into sub-themes (Figure 4.6).
Separate chapters are devoted to each of the major themes. The use of direct quotations
has been used to explore the sub-themes in the following results chapters, as they are:
97
A basic source of raw data in qualitative inquiry, revealing respondents’ depth of
emotion, the ways they have organised their world, their thoughts about what is
happening, their experiences, and their basic perceptions (Patton, 2002, p. 21).
Figure 4.6: Coding tree
4.5 Quantitative component
This study had some secondary questions, and in order to answer the questions, ‘Where
do Snowy Monaro residents die?’, ‘What factors influence the PoD?’, and ‘Are
residents living on farms less likely to die at home than residents living in town?’, the
inclusion of a survey method was necessary. Ethics approval was granted as previously
reported.
Quality rural
end-of-life care
Family factors
Patient factors
Importance of place
The ‘good death’
Rural
Place
Formal support
Informal support
Family
Benefits and
challenges
RACF - home or
institution
Not home
Importance of home
Meaning of home
98
4.5.1 Quantitative methods objectives
A formal quantitative approach allowed me to describe the PoD of residents within the
Snowy Monaro region, both systematically and explicitly. It also allowed me to present
the data visually (charting) and mathematically (frequencies and logistic regression) in
order to identify patterns within, and relationships between the influencing factors
associated with dying at home or in one’s usual place of residence (UPoR). These
patterns may not be obvious from the qualitative interviews, due to the small sample
size. Death data were collected from a variety of local informants, including funeral
directors, residential aged care and inpatient health facilities, local newspaper obituary
and funeral notices, and funeral announcements on local radio. An explanation of the
data collection and analysis is provided later in this chapter.
4.5.2 Recruitment
The final group of participants in this original research, were the informants of the
population PoD survey data. The primary informant was the local funeral director. I had
approached him, prior to gaining HREC approval, to seek his interest in participating.
Once ethics approval was obtained, he was formally invited to participate and was
issued with an information sheet and consent form, (copy in Appendix 4). He was
advised that participation was voluntary, and that he could withdraw at any time, until
publication of the data. It was made quite clear to the funeral director that, as the only
‘local’ funeral director, his identity would easily be recognised by anyone familiar with
the region and reading this thesis or future journal publications.
As data collection began, it became apparent that data relating to decedents residing in
the Bombala area were missing, due to adjacent out-of-region funeral directors being
engaged, and lack of obituaries published in the once weekly ‘Bombala Times’. An
amendment was made to the ethics approval to recruit additional informants. As there
are no resident GPs in the town, most deaths in the Bombala district, regardless of PoD,
are certified at the local hospital (Bombala MPS), before the body is transported to the
undertakers’ mortuary. There was also conflicting information, provided in the obituary
notices, on some long term RACF residents, with the obituaries reporting their last
address as their previous private residence rather than the RACF. This fact was
highlighted during a conversation with one DON. A further amendment was made to
the protocol to permit DONs to provide the length of stay for their deceased residents,
99
identified through obituary and funeral notices. Contact with the informants was in
person or via email or phone.
Waiver of consent from decedents’ next of kin was granted on the basis that names and
addresses were not collected, and the use of any data was unidentified. The majority of
decedents had previously been identified either in local newspaper obituary notices or
funeral announcements on local radio.
4.5.3 Quantitative data collection
Death data were collected from 1st February 2015 through 31st May 2016. Denominator
data on numbers of deaths were obtained from consenting informants (local funeral
director (personal contact and website), Bombala MPS and RACFs), and from
obituaries and funeral notices in local newspapers, on local radio, and adjacent out-of-
region funeral directors’ websites (Table 4.3) After contacting the NSW Registry of
Births, Deaths and Marriages by email, to seek advice as to how best to access data
contained in the official death certificates, I was referred onto the Australian Bureau of
Statistics (ABS). The ABS advised that their Causes of Death dataset does not contain
the actual place of death beyond the state or territory the death was registered. For
ethical reasons, mainly relating to privacy and consent, I was unable to personally
access copies of the Medical Certificate Cause of Death (MCCD) held by the local
hospitals, funeral directors, RACFs and GPs. Cause of death was therefore determined
by data provided by the consenting informants taken from their official records. De-
identified data on date of death, age at death, sex, place of death, cause of death, and last
place of residence (farm, village, or town) was supplied by the informants. RACFs also
provided decedents’ length of stay. Duplicate records were deleted by cross checking
obituary and funeral notices with data received from the consenting informants.
It was my intention to collect data over a 2-year period. At the start of data collection
there was a widely circulated, twice weekly local newspaper, the ‘Monaro Express’,
publishing obituary and death notices, from which data on ‘last place of residence’ and
marital status were collected. Without notice, the newspaper ceased publication in mid-
May 2016. At the time, many obituary and funeral notices were not published in the
remaining weekly local paper “The Monaro Post”, due to its small circulation, and
distribution after many funerals had taken place.
100
Table 4.3: Informants of the Snowy Monaro death data
Informant: Total deaths, N=224 n (%)
Local funeral director and local newspaper
Local funeral director (direct and website)
Adjacent out-of-region funeral directors’ websites
Local newspapers (no other informant)
Local radio only (no other informant)
BMPS nursing manager or RACF DON
Qualitative bereavement interview
122 (54.5%)
48* (21%)
8 (4%)
25 (11%)
2 (1%)
18 (8%)
1 (0.5%)
*Includes 4 deaths identified only on website; BMPS- Bombala multi-
purpose service; RACF residential aged care, DON director of nursing.
161 deaths (72%) were available in the public domain.
4.5.4 Quantitative analysis
The population PoD data was entered into IBM Statistical Package for Social Sciences
version 22 (SPSS-22), for ease of recording and analysis. Demographic data were
collected on date of death, sex, age at death, marital status, last place of residence, PoD,
and cause of death. Marital status was only available from newspapers, resulting in
missing data for 31.7% of decedents. Marital status was classified into married/de facto/
widowed/single/unknown; for the logistic regression model, marital status was
dichotomised into partnered (married/de facto) and un-partnered (widowed/single). It
was not possible to determine if decedents were living alone or with others, or how long
patients were cared for at home before transferring to a hospital or RACF.
Place of residence was categorised according to the degree of rurality, which reflected
the availability of health services. Only the towns of Cooma and Bombala were
classified as ‘town’, as they contained GPs, rural hospitals and RACFs. The remaining
region was classified as ‘small towns’ (with GPs), ‘villages’ (no GPs); or ‘farm’. Cause
of death was separated into ‘natural’ or ‘external’ (accident, suicide, homicide) causes.
Of those deaths considered to be ‘natural’, decedents were dichotomised into ‘palliative’
or ‘non-palliative’. The palliative subgroup consisted of decedents who had died of an
advanced chronic condition. These included advanced frailty (Gomez-Batiste et al.,
2014), and the 10 conditions, considered by McNamara et al. (2006) to be amenable to
palliative care (cancer (solid and haematological), chronic cardiac failure (CCF), end
stage renal failure (ESRF), liver failure, chronic obstructive pulmonary disease (COPD),
motor neurone disease (MND), Parkinson’s disease, Huntington’s disease, dementia,
and HIV/AIDS). In this study, there were no deaths due to Huntington’s disease or
101
HIV/AIDS. While other studies (Gomez-Batiste, ibid) have included chronic cerebral-
vascular disease in the palliative subgroup, as it was uncertain if deaths due to
cerebrovascular accident (CVA) were acute or chronic they were excluded from the
palliative subgroup. It is unknown which, if any, of the decedents in the palliative
subgroup were identified by their health professionals as ‘palliative’ or if they were
receiving palliative care at the time of their death.
Data were initially charted descriptively. Cross tabulations by PoD and key
demographic data were performed. Bivariate logistic regression was undertaken to
identify factors that may influence the actual PoD, in particular dying in one’s UPoR,
and within region. Independent variables used in linear regression modelling were
derived from studies reported in my literature review, and consistent with factors
emerging from the original research qualitative data analysis.
The outcome variable was place of death: private residence, local hospital, out-of-region
hospital, out-of-region (Canberra) inpatient hospice, RACF, other (road, workplace),
and unknown. RACF decedents fell into two groups: those who were permanent
residents (length of stay (LOS) ≥ 3months), for whom the RACF was their ‘usual place
of residence’ (UPoR); and those admitted for respite or transferred from their private
residence for end-of-life care (LOS < 3 months). For this study UPoR referred to either
a private residence or long term RACF. The final dependent variables were ‘dying in
UPoR’ or ‘not dying in UPoR’, and ‘dying in region’ or ‘dying out of region’.
Pearson Chi2 tests were used to determine if any of the independent variables were
associated with PoD. Exact p-values were calculated whenever there were any expected
cell counts less than five. For large contingency tables, Monte Carlo simulation was
used to estimate exact p-values. P-values less than 0.05 were considered to be
statistically significant. Bivariate logistic regression was conducted using the model
outlined in figure 4.7. Multivariate logistic regression was unachievable.
102
Figure 4.7: SPSS modelling for logistic regression
Abbreviations: SPSS=Statistical program for social sciences; UPoR=usual place of residence
4.6 Ethical considerations
Both qualitative and quantitative research received ethics approval from two human
research ethics committees, the Australian National University (ANU) and Greater
Western New South Wales Local Health District (GWNSW LHD) Human Research
Ethics Committees. Site specific approval was also granted by the Southern New South
Wales Local Health District (SNSW LHD) to conduct the research within SNSW LHD
facilities, and to interview their staff and patients. A copy of the ethics approvals can be
found in Appendix 1.
4.7 The role of the researcher
Research does not occur within a vacuum, instead the researcher brings to the
study/interview their own insights, opinions and beliefs formulated from personal
experience. To maintain rigor within the study the researcher must remain reflective:
The qualitative analyser owns and is reflective about [their] own voice and perspectives;
a credible voice conveys authenticity and trustworthiness; complete objectivity being
impossible and pure subjectivity undermining credibility, the researcher’s focus
Analyse
Generalised
linear
models
• UPoR (lowest value)
Binary
logistics
• UPoR/Not UPoR
• Died in region/out of region
Dependent variable
Reference
Predictors • Sex, age, marrital, cause of death, last residence
103
becomes balance - understanding and depicting the world authentically in all its
complexity while being self-analytical, politically aware, and reflexive in consciousness
(Patton, 2002, p. 41).
Throughout each step of this study I have tried to be transparent and as objective as
possible; however, the circumstances under which this study was conducted were
inherently open to bias. Firstly, I am the only palliative medicine specialist in the region
and a long term resident, over 25 years. Secondly, one of my supervisors is also a long
term resident and medical practitioner in the Snowy Monaro region. As such, we are
well known and (hopefully) respected. This had potential to create bias in recruitment
and a sense of participant coercion. As previously noted, a third party conducted the
consent process for any of my current patients. A few patients declined the recruiters’
invitation to participate. In contrast, Gwen, whose story is told in the preface to Chapter
Six, was extremely keen to participate. Gwen had metastatic cancer. Living in a country
town meant I was not just her GP, but our paths also crossed socially. Long before I had
gained ethics approval for this study, Gwen had heard of my project and asked if there
was anything she could do to help me with the study. She offered to be interviewed, and
wanted to “be a help. That’s the main thing”. While waiting for ethics approval, Gwen
asked on numerous occasions, “when are you going to interview me?” When the green
light was finally given, Gwen was participant number one. To appease the ethics
committees’ fear that recruitment could be seen as coercion, I arranged for our practice
nurse to complete the consent process.
During interviews there was an imbalance of power: interviewer/doctor versus
interviewee/patient. I tried to be transparent and trustworthy in conducting the
interviews by reinforcing the fact that during the interviews I was a researcher and not a
doctor, and would not be making any comment or giving advice on management. The
ethics committees also clearly stated that I could not use any information I had prior
knowledge of, or that had not been obtained in interviews. This meant that I could not
access the medical notes, of not only the interviewees but also to fill in missing data
collected for the death statistics component of the study. This was a two way restriction,
and in one case I was unable to share potentially significant information I had gathered
from a confidential patient interview, with family and health providers during a
management related family meeting.
104
While I was involved in the management of only three patients at the time of
recruitment and interview, as circumstances dictated I was actively involved in the
management of an additional five patients prior to their death. While this may not have
influenced the collection of data, there is potential for my knowledge of the timing and
circumstances of each one’s death to influence my analysis and write up of my report,
because, as one university lecturer told me, “You cannot un-know what you already
know” (Dr Graham Fordham, Senior Lecturer in Social Foundations Medicine, Medical
School, ANU College of Medicine, Biology and Environment, 2015). Likewise, of the
three previously bereaved relatives, I had been involved in the care of two of their
family members.
4.8 Limitations
The limitations of this study are described in detail in the final chapter of this thesis.
However, I believe it is fitting to acknowledge at this point, limitations that were
identified early in the development stage of this research, as a direct result of my
conflict of interest. As a medical practitioner, my primary aim is to ensure symptoms
are managed, and that patients are comfortable and settled. Therefore, there is potential
risk of my pre-conceptions influencing the findings. In Chapter One, introduction to this
thesis, I talked about my own personal and professional experiences with dying persons,
and that my belief is that home is not necessarily the best PoD. While some patients and
their FCGs will compromise ‘best care’ and comfort to remain home, this is not my
value. Escalating pain is more readily addressed in hospital than at home, especially out
on farms. I have tried to put my beliefs aside, listen to, and honestly portray the views
expressed by the rural patients and their FCGs. For that reason, I have included long
quotes and narratives in Chapters Six, Seven and Eight.
4.9 Rigour of the research
The central focus of this study was on finding meaning behind the quantitative data
describing the actual PoD within the Snowy Monaro region. This was undertaken using
an ethnographic approach. In many ways, qualitative research has polarised the
scientific world, over issues such as validity, reliability and objectivity (Patton, 2002,
105
p. 93). Even qualitative researchers are polarised into those who ‘emphasise the
importance of rigorous qualitative research and those who regard “rigour” as
inappropriate’ (Grbich, 1999, p. 61). One of the criticisms of qualitative research is that
the analysis is subjective and influenced by researcher bias. Kuper et al. (2008) suggest
the most important component for the reader to assess the rigour and relevance of
qualitative research is to have a strong understanding of the methodological approach
and methods utilised by the researcher. I am quietly confident that, after reading my
literature review, the previous chapter describing the setting, and this chapter explaining
the methodological approach and methods for collecting and analysing data, the reader
will be confident that the results of this study will be valid, reliable and objective. I have
attempted to systematically approach all aspects of this research.
While it is important that all work I present in this thesis is substantially my own, in
order to satisfy the rigorous standards required for systematic and scoping literature
reviews, my supervisors willingly assisted in the validation of the included studies. One
supervisor independently assessed all the titles of identified papers. All my supervisors
independently reviewed a portion of the retained abstracts. However, I conducted the
literature searches and independently undertook every step of the literature review
process.
Sampling strategies for the qualitative component were appropriate for the study
objectives. While the focus of this research was on patients with a life limiting illness
and their FCGs, in order to gain perspectives on the ‘good’ rural death through different
lenses, the HCPs were also interviewed (also known as triangulation). This allowed
comparison and corroboration of data. Recruitment did not initially use a sampling
frame, but as recruitment progressed it became purposive, seeking to incorporate a
diverse range of participants. Data collection was over a period of time, with a few
participants being interviewed more than once. In order to ensure authenticity in
reporting, I have provided quotes to represent the participants’ views and perspectives.
As reported in the preceding section I have, to the best of my ability, honestly and
transparently acknowledged by bias in all components of this research.
Rigour in the collection, analysis and reporting of the quantitative data was enhanced by
collecting whole population data over a 16 month period. When it became apparent that
106
much of Bombala Shire data were missing from the records of the local funeral director,
the protocol was amended in an attempt to capture as much data as possible. Adoption
of a mixed methods approach meant that PoD data obtained in the bereavement
interviews could be validated.
This chapter has presented the theoretical framework underpinning my research. The
methods used to collect and analyse data have been described, and the participants
introduced. This completes Part 2 of the thesis, the report on the literature review,
preparatory, and data collection and analysis phases of the research study. The results of
this research study are reported, discussed and argued in Part 3, consisting of four
chapters. The quantitative data are presented in the next chapter (Chapter Five). The
qualitative data collected from the interviews are then presented separately, in the three
chapters that follow (Chapters Six, Seven and Eight).
107
PART 3: RESEARCH STUDY RESULTS
108
109
Introduction to Part 3: Research study results
Part 3 of this thesis reports the quantitative results and qualitative findings. The concept
of the good death, place of death (PoD), and rural residency, are discussed, as described
by the participating patients, family caregivers, and healthcare professionals. Four major
themes emerged, with a chapter devoted to each theme.
Chapter Five provides the context for the qualitative data, by describing the actual PoD
of Snowy Monaro residents, between February 1, 2015 and May 31st, 2016. These data
tell us where the residents died, but do not provide any insight into the reasons that
determined each PoD.
Chapter Six discusses the meaning and importance of place of care and death, from the
experiences and perspectives of the participating patients, and their FCGs and rural
HCPs. The primary objective of this thesis was to explore the concept of the good death,
therefore, Chapter Seven is allocated to the participants’ narrative on the ‘good rural
death’, including their perspectives on the influence of ‘place’ on the ‘good death’.
Chapter Eight discusses the influence of rural residency on the PoD and the concept of
the good death.
Each of the three qualitative results chapters (Chapters Six, Seven, and Eight) begins
with a vignette, the story of one participating patient. This sets the scene for what is
discussed in each chapter. The issues raised by each highlighted patient are woven
through the remainder of the chapter. The body of each chapter contains long narratives,
provided by the participants. As a palliative medicine specialist and researcher, my
interpretations are open to my biases. To maintain transparency, it is important for the
reader to ‘hear’ from the participants, rather than from my paraphrasing. Rather than
numbers, aliases (names and ages have been changed) have been used to introduce the
participants, as they are real people, and valued members of my community.
Part 3 begins by describing the actual PoD of Snowy Monaro residents.
110
111
CHAPTER FIVE: PLACE OF DEATH OF SNOWY
MONARO RESIDENTS
Introduction
This chapter describes the actual place of death (PoD) of Snowy Monaro residents,
between February 1, 2015 and May 31, 2016. The chapter begins by describing the
socio-demographic features, cause of death, and PoD for the full data set (Section 5.1).
While the focus of this thesis is the palliative subgroup, Section 5.2 briefly describes the
cause of death and PoD for the non-palliative subgroup, i.e. those who died
unexpectedly of an acute condition. Section 5.3 describes the data related to those
residents who died of an advanced chronic condition, considered amenable to palliative
care. The results specific to RACF deaths are reported and discussed in Section 5.4.
5.1 Description of all deaths
Over the 16-month data collection period, death records of 224 Snowy Monaro
residents were collected. One hundred and ninety (85%) died as a result of ‘natural’
causes, and five (2%) of ‘external’ causes. The external causes of death included one
death from injuries sustained in a motor vehicle accident, one suicide, one drug
overdose, one homicide, and one death in a house fire. Natural deaths included 52
(27%) from an acute or ‘non-palliative’ condition and 138 (73%) from a ‘palliative’
illness. The cause of death was unknown for 29 (13%) residents.
Table 5.1 describes demographic features and PoD of decedents. There were no
significant differences between the palliative and non-palliative groups in relation to
their likelihood of dying in their usual place of residence (UPoR) or dying in the region.
Except for one neonatal death (excluded from the analysis), the age at death ranged
from 20 to 101 years. Four residents were 100 years or older. Most decedents (66%)
lived in Cooma or Bombala, i.e. one of the two towns with a hospital. The 66 deaths in
RACFs comprised two groups: 51 permanent residents (77%) and 15 short stay
residents [those admitted for either respite, or transferred from their private residence or
hospital for end-of-life care (23%)].
112
Table 5.1: Decedents’ demographics and place of death
All deaths n=224# Palliative
n=138 (62% *)
Non-palliative
n=52 (23%*)
Sex
Male 116 (52%) 75 (54%) 23 (44%)
Female 108 (48%) 63 (46%) 29 (56%)
Age (years)
≤39 3 (1%) 0 1 (2%)
40-49 6 (2.5%) 4 (3%) 1 (2%)
50-59 13 (6%) 9 (6.5%) 2 (4%)
60-69 31 (14%) 20 (14.5%) 5 (9.5%)
70-79 53 (24%) 37 (27%) 9 (17%)
80-89 69 (31%) 43 (31%) 19 (36.5%)
≥90 46 (20.5%) 24 (17%) 15 (29%)
unknown 3 (1%) 1 (1%) 0
Marital
Partnered+ 82 (36%) 65 (47%) 10 (20%)
Unpartneredǂ 71 (32%) 43 (31%) 21 (40%)
Unknown 71 (32%) 30 (22%) 21 (40%)
Place of residence
Farm 33 (14.5%) 24 (17%) 2 (4%)
Village (no GP) 18 (8%) 10 (7%) 5 (10%)
Small town (GP, no hospital) 17 (7.5%) 8 (6%) 6 (11%)
Town with GP and hospital 147 (66%) 91 (66%) 39 (75%)
Unknown 9 (4%) 5 (4%) 0
Place of death
Private residence 42 (19%) 17 (12%) 14 (27%)
Local hospital 82 (36.5%) 64 (46%) 17 (32.5%)
RACF ## 66 (29.5%) 44 (32%) 14 (27%)
Tertiary hospital 16 (7%) 8 (6%) 5 (9.5%)
Hospice (out of region) 4 (2%) 4 (3%) 0
Other place 5 (2%) 1 (1%) 2 (4%)
Missing data /unknown 9 (4%) 0 0
Died in UPoR p= 0.41♦
Yes 93 (41.5%) 52 (38%) 23 (44%)
No 122 (54.5%) 86 (62%) 29 (56%)
Unknown 9 (4%) 0 0
Died within the region p=0.89♦
Yes 191 (85.5%) 122 (88.5%) 47 (90%)
No 21 (9.5%) 14 (10%) 5 (10%)
Unknown 12 (5%) 2 (1.5%) 0
# ‘All deaths’ includes unknown cause of death (n=29;13%) and external causes (n=5, 2.2%); * percentage of
all deaths; + Partnered= married/de facto; ǂ Unpartnered= single, widowed, divorced;
##RACF= residential aged care facility, includes permanent and respite/short stay residents; GP=general
practitioner; UPoR= usual place of residence, includes private residence and permanent placement ≥ 3months
in RACF; %= column percentages. ♦P values for dying in UPoR and within region >0.05 (by Pearson 2 test
for differences in proportions between palliative and non-palliative groups;2x2 table, unknown excluded)
113
5.2 Non-palliative subgroup
Of the 190 natural deaths, 52 (23% of all deaths) were considered to be due to a non-
palliative condition (Figure 5.1). Of these, half were due to a cardiovascular cause
(cardiac arrest or acute myocardial infarct). The remaining causes were cerebrovascular
accident (CVA), pneumonia, sepsis, subarachnoid haemorrhage (SAH), and five of
another cause (abdominal aortic aneurysm, seizure (epilepsy), fractured neck of femur,
bowel obstruction and aspiration). Where cause of death was reported to be due to a
CVA, the decedent was classified as non-palliative, as it was not possible to determine
if they died of an acute event. It is therefore possible that some CVA survivors may
have been incorrectly classified and could have been considered ‘palliative’.
Place of death for the non-palliative subset was evenly distributed between private
residence, local hospital and RACF. Just under ten percent died in the out-of-area
tertiary hospital in Canberra. As the local hospitals are staffed by general practitioners,
with no intensive care, cardiac care, interventionists, specialist anaesthetists or onsite
surgeons, serious, unexpected acute conditions are transferred to the Canberra Hospital.
26
12
4
3
2
5
Figure 5.1 Cause of death of the non-palliative subgroup, n=52
Cardiac CVA Pneumonia Sepsis SAH Other
114
5.3 Palliative subgroup
The focus of this thesis is the palliative subgroup, consisting of decedents reported to
have died of an advanced chronic condition considered amenable to, but not necessarily
receiving, palliative care. There were 138 (62% of all deaths) in the palliative subgroup.
In this study, the causes of death amenable to palliative care included cancer (solid and
haematological), chronic cardiac failure (CCF), end stage renal failure (ESRF), liver
failure, chronic obstructive pulmonary disease (COPD), motor neurone disease (MND),
dementia (McNamara et al., 2006), and advanced frailty (Gomez-Batiste et al., 2014).
Full details of the statistical methods are found in Section 4.7.1 of the Methods Chapter.
Cancer was reported as the cause of death for 66 (48%) of the palliative deaths. cancer
Seventy-two (52%) non-cancer causes of death included chronic obstructive pulmonary
disease (COPD) or pulmonary fibrosis; chronic cardiac failure (CCF); end stage renal
failure (ESRF); and ‘other’. The ‘other’ category included liver failure (five) and motor
neurone disease (one) (Figure 5.2).
In the palliative subgroup, ages ranged from 40 years to 100 years; median 79 years.
Consistent with my literature review, most patients who died of a condition amenable to
palliative care, died in a place away from their private residence. The ‘other’ category
66
12
21
25
44 6
Figure 5.2 Deaths caused by a condition amenable to palliative care, n=138
Cancer
COPD/Pulmonary fibrosis
Chronic cardiac failure
Dementia
End stage renal failure
Fraility
Other
115
related to one resident with chronic heart failure who died while staying at a community
respite cottage.
Of the 138 palliative deaths, 17 (12%) occurred at ‘home’ in a private residence and 121
(88%) occurred in an institution (hospital, hospice or RACF). In regards the out-of-
region hospital, eight died in the tertiary hospital, 115 kilometres from Cooma. There
were no reported deaths in out-of-region private hospitals, Bega Regional Hospital, or
other tertiary hospitals.
5.3.1 Place of death according to cause of death
In this sample, cause of death included cancer and non-cancer diagnoses. Deaths due to
COPD, CCF, ESRF, dementia, frailty, liver failure and MND were recorded in the
advanced non-cancer chronic conditions amenable to palliative care category. In Table
5.2 the ‘other’ category includes liver failure and MND. There were no deaths due to
Parkinson’s disease, Huntington’s disease or HIV/AIDS.
Table 5.2 describes the PoD, by socio-demographic features, for the palliative subgroup.
Nearly half of all palliative deaths occurred in the local hospital, one-third in a RACF,
one-in-eight in a private residence. People who died of cancer were more likely to die in
the local hospital, while people who died of non-cancer causes were more likely to die
in a RACF (Monte Carlo simulation exact p=0.001). Of the non-cancer causes of death,
people with dementia were the most likely to die in a RACF (Monte Carlo simulation
exact p= 0.04). No residents over the age of 70 years or with a non-cancer diagnosis
died in the out-of-region hospice. Across the palliative subgroup, 125 (91%) died within
the region.
5.3.2 Place of death according to degree of rurality
No deaths occurred in a village private residence. The results suggested that people
living on farms, and in villages and small towns were more likely to die in hospital.
People living in Cooma or Bombala were more likely to die in a RACF or the local
hospital. However, when the Monte Carlo simulation test was applied, no statistical
differences were found. Three farmers died in a RACF; however, it was not possible to
determine how many farmers had moved permanently to a RACF (Table 5.2).
116
Table 5.2: Place of death for the palliative subgroup by socio-demographic characteristics. N=138
Total
number of.
deaths (% of
total)
N (%)
occurring at a
private
residence
N (%)
occurring at a
local hospital
N (%)
occurring at
an out-of-
region
hospital
N (%)
occurring at
the out-of-
region
hospice
N (%)
occurring in a
RACF *
N (%)
occurring at
another place
p-value♦
All palliative deaths 138 17 (12%) 64 (46%) 8 (6%) 4 (3%) 44** (32%) 1 (1%)
Sex
Male 75 (54%) 12 (16%) 38 (51%) 4 (5%) - 20 (27%) 1 (1%) 0.069
Female 63 (46%) 5 (8%) 26 (42%) 4 (6%) 4 (6%) 24 (38%) -
Age (years)
40-49 4 (3%) 1 (25%) 1 (25%) - 1 (25%) 1 (25%) - 0.039
50-59 9 (6.5%) 4 (44.5%) 3 (33.5%) 1 (11%) 1 (11%) - -
60-69 20 (14.5%) 4 (20%) 12 (60%) 2 (10%) 1 (5%) 1 (5%) -
70-79 37 (27%) 7 (19%) 18 (48%) 3 (8%) 1 (3%) 7 (19%) 1 (3%)
80-89 43 (31%) 1 (2%) 19 (44%) 2 (5%) - 21 (49%) -
≥90 24 (17%) - 10 (42%) - - 14 (58%) -
unknown 1 (1%) - 1 (100%) - - - -
Marital status
Partnered+ 65 (47%) 9 (14%) 33 (51%) 4 (6%) 2 (3%) 16 (25%) 1 (1%) 0.315
Unpartneredǂ 43 (31%) 4 (9%) 17 (39.5%) 1 (2.5%) 2 (4.5%) 19 (44%) -
Unknown 30 (22%) 4 (13%) 14 (47%) 3 (10%) - 9 (30%) -
117
Table 5.2 continued
Total N Private
residence
Local hospital Out-of-region
hospital
Out-of-region
hospice
RACF * Another place p-value
Residence 0.126
Farm 24 (17%) 5 (21%) 12 (50%) 3 (12.5%) 1 (4%) 3 (12.5%) -
Village (no GP) 10 (7%) - 8 (80%) - - 2 (20%) -
Small town (GP,
no hospital)
8 (6%) 3 (37.5%) 5 (62.5%) - - - -
Town (GPs+
hospital)
91 (66%) 9 (10%) 35 (38.5%) 5 (5.5%) 3 (3%) 38 (42%) 1 (1 %)
Unknown 5 (4%) - 4 (80%) - - 1 (20%) -
Cause of death
Cancer 66 (48%) 9 (14%) 38 (57.5%) 5 (7.5%) 4 (6%) 10 (15%) - 0.001
Non-cancer
72 (52%) 8 (11%) 26 (36%) 3 (4%) - 34 (47%) 1 (2%)
Non-cancer N=72 n=8 n=26 n=3 n=0 n=34 n=1 0.04
COPD 12 (17%) 3 (25%) 5 (41%) 2 (17%) - 2 (17%) -
CCF 21 (29%) 1 (5%) 10 (47%) 1 (5%) - 8 (38%) 1 (5%)
Dementia 25 (3%) - 4 (16%) - - 21 (84%) -
ESRF 4 (5.5%) 1 (25%) 2 (50%) - - 1 (25%) -
Frailty 4 (5.5%) 1 (25%) 2 (50%) - 1 (25%) -
Other§ 6 (8%) 2 (33%) 3 (50%) - - 1 (17%) -
Abbreviations: *RACF= residential aged care facility (includes both permanent (35) and short-stay residents (9)); GP=general practitioner; COPD= chronic
obstructive pulmonary disease; CCF= chronic cardiac failure; ESRF= end stage renal failure. Presented percentages are row percentages except for Column 2 “total
number of deaths’ (column percentages). ** 1 resident died in an out of region RACF. § Other= 5 liver failure and 1 motor neurone disease, (no deaths due to
Parkinson’s disease, Huntington’s disease, or HIV/AIDs; + Partnered= married/de facto; ǂ Unpartnered= single, widowed, divorced.
♦ Monte Carlo simulation test, exact p-values less than 0.05 indicate an association between place of death and the factor of interest.
118
5.3.3 Deaths in usual place of residence
As previously noted, PoD was initially separated into six categories, with RACF deaths
including those of both permanent residents and those admitted for respite or transferred
from home or hospital for end-of-life care. As will be discussed in Chapter Six, some
long-term residents of RACFs considered the facility to be their ‘home’ and not an
institution. Therefore, the category ‘private residence’ does not represent all the deaths
that occurred at ‘home’. When permanent RACF residents’ deaths in their facility were
combined with deaths in a ‘private residence’, the number of deaths occurring within an
‘institution’ decreased from 88% to 62%. The term ‘usual place of residence’ (UPoR)
included private residences and RACFs for permanent residents. Thirty-five (49%) of
non-cancer and 17 (26%) of cancer deaths occurred in an UPoR. The following chapter
discusses the rural participants’ experiences and perspectives on PoD, including
RACFs.
5.3.4 Factors influencing death at usual place of residence
The binary logistic regression model found no statistically significant associations
between sex, age, or marital status and dying in one’s UPoR (Table 5.3). The bivariate
logistic regression showed that people with a non-cancer diagnosis were more likely to
die in their UPoR than those who died from cancer (odds ratio (OR) 2.73, 95% CI 1.33,
5.60; p=0.005), with 84% of those with dementia dying in a RACF.
Figure 5.3 charts the percentage of deaths occurring in UPoR according to cause of
death, and the error bars indicating the 95% Wald Confidence Interval. Those living on
farms and in small towns were less likely to die in their UPoR than town dwellers (OR
0.31, 95% CI 0.11, 0.89; OR 0.31, 95% CI 0.10, 1.01 respectively; p 0.016).
26
3329
8050
25 33
0
20
40
60
80
100
Cause of death
Per
centa
ge
of
dea
ths
in U
Po
R.
Figure 5.3 Deaths in usual place of residence according to cause of death
for the palliative subgroup.
Cancer COPD CCF Dementia Frailty ESRD Other
119
5.3.5 Factors influencing death within region
A binary logistic regression model using ‘dying within region’ as the outcome variable
found no statistically significant associations between sex, age, marital status, or cause
of death, and dying within region. The model suggested that farm dwellers were less
likely to die within region than town residents (OR 0.55 95% CI 0.15, 1.96). No small
town residents died out of region; therefore, it was not possible to estimate the odds
ratio for small town residents dying within region (see ‘not estimable’ (NE) Table 5.4).
Table 5.3: Odds ratio estimates and 95% confidence limits of factors associated with
dying in usual place of residence and in region, for the palliative subgroup. Dying in usual place of residence Dying within region
OR 95% CI OR 95% CI
Sex
Male (n=75) 0.86 0.43, 1.7 2.04 0.63, 6.57
Female (n=63) 1 1
Age (years)p
≤64 (n=22) 0.53 0.18, 1.52 0.22 0.05,1.03
65-84 (n=66) 0.57 0.26, 1.21 0.79 0.18, 3.50
≥85 (n=49) 1 1
Marital
Unpartnered (n=43)ǂ 1.51 0.68, 3.35 0.99 0.26, 3.74
Partnered (n=65)+ 1 1
Cause of Death§p
ESRF (n=4) 0.96 0.09, 9.87 NE NE
COPD (n=12) 1.44 0.39, 5.40 0.79 0.15, 4.21
Dementia n=25 11.53 3.72, 35.50 3.79 0.46, 31.58
Frailty (n=4) 2.88 0.38, 22.08 NE NE
CCF (n=21) 1.15 0.39, 3.45 3.16 0.38, 26.52
Cancer (n=66) 1 1
Palliative typep
Non-cancer (n=72) 2.73 1.33, 5.60 2.68 0.79, 9.18
Cancer (n=66) 1 1
Residencea
Farm (n=24) 0.31 0.11. 0.89 0.55 0.15, 1.96
Small town (n=19)♦ 0.31 0.10, 1.01 NE NE
Town (n=91) 1 1
Abbreviations: OR=Odds ratio; CI= confidence interval; ESRF=end stage renal failure; COPD= chronic
obstructive pulmonary disease; CCF=chronic cardiac failure; NE= not estimable; + Partnered=
married/de facto; ǂ Unpartnered= single, widowed, divorced; § excludes ‘other’ causes; ♦Small towns
with GPs/no hospital; p= p-value <0.05, indicating an association between dying in UPoR and factor of
interest. First column excludes ‘unknown’ (see table 5.2 for percentages of missing data).
120
5.4 Deaths in residential aged care facilities
During the data collection period, 65 decedents (29% of all deaths) were identified as
permanent RACF residents (median length of stay 25 months, range four -168 months). Ages
ranged from 69 to 101 years, median 88 years. Of these residents, 51 (78.5%) died at the
facility (Table 5.4). In total, 66 deaths occurred in a RACF, including eight (12%) whose cause
of death was not available. The length of stay before death, for the 15 admitted for respite or
EoLC, ranged from one day to just short of three months.
One RAC resident died in the out-of-region tertiary hospital (cause of death not available as
referred to the Coroner). Nearly 30 percent of the palliative subgroup within the Snowy
Monaro region died in a RACF, and more than one third were due to dementia (Table 5.4).
Dementia is rapidly becoming one of the main causes of death in Australia (ABS, 2015),
with many dementia sufferers residing, and therefore dying, in RACFs (Black et al., 2016).
Table 5.4: Characteristics and PoD for all permanent RACF residents, n=65.
Total n=65 (29% of total deaths)
Sex
Male 27 (41.5%)
Female 38 (58.5%)
Age (years)
60-69 1 (1.5%)
70-79 8 (12.5%)
80-89 30 (46%)
≥90 26 (40%)
Marital status
Married 16 (24.5%)
Widowed 27 (41.5%)
Unknown 22 (34%)
Place of death
RACF 51 (78.5%)
Local hospital 13 (20%)
Tertiary hospital 1 (1.5%)
Died in UPoR
Yes 51 (78.5%)
No 14 (21.5%)
Type of death
Palliative 42 (65%)
Acute/non-palliative 15 (23%)
Unknown 8 (12%)
Palliative deaths n=42
Cancer 8 (19%)
Non-cancer 34 (81%)
121
Table 5.4 continued Total n=65
Non-cancer palliative death n=34
COPD 2 (6%)
Dementia 22 (65%)
Frailty 1 (3%)
CCF 8 (23.5%)
Other 1 (3%)
Non-palliative death n=15
Cardiac (AMI) 4 (27%)
CVA 3 (20%)
Sepsis/pneumonia 5 (33%)
Other 3 (20%)
Abbreviations: PoD= Place of death; RACF=residential aged care facility; UPoR= usual place of
residence; COPD=chronic obstructive pulmonary disease; CCF= chronic cardiac failure; AMI= acute
myocardial infarction; CVA=cerebrovascular accident (onset unknown).
Non-palliative cause of death: Other= bowel obstruction, ruptured aortic aneurysm, fractured femur.
5.5 Palliative deaths, excluding permanent RACF residents
During the data collection period, more Snowy Monaro residents died in hospital
(43.5%) than at home (19%), in their private residence, reflecting the general trend
noted in the literature for rural patients to die in hospital rather than home. One
significant inconsistency in the literature is the definition of ‘home’ and ‘institution’,
especially when permanent RACF residents have been included in the reported findings.
While not specifically a rural study, McNamara and Rosenwax (2007) dichotomised
PoD into ‘UPoR’ (private residence and permanent placement in RACFs), and ‘place
other than UPoR’. However, most previous studies have considered RACFs to be
‘institutions’ and therefore, not ‘home’.
While the aim of the quantitative component of this study was to determine where
Snowy Monaro residents die, the large number of permanent RACF residents who
mostly reside in one of the two towns, and die in their UPoR, is one of the reasons for
the preponderance of town residents not dying in hospital. A secondary analysis of the
data was undertaken on the palliative subgroup, by excluding permanent RACF
residents, to determine where non-RACF residents die. (Table 5.5).
122
Table 5.5: Socio-demographics of palliative sub-groups: including, excluding, and exclusive of
permanent residential aged care facility residents.
All Palliative
N=138*
Palliative subgroup,
excluding
permanent RACF
residents; n=92
Palliative subgroup:
permanent RACF
residents only; n=42
Sex
Male 75 (54%) 57 (62%) 16 (38%)
Female 63 (46%) 35 (38%) 26 (62%)
Age (years)
40-49 4 (3%) 4 (4%) 0
50-59 9 (6.5%) 9 (10%) 0
60-69 20 (14.5%) 18 (19.5%) 0
70-79 37 (27%) 31 (34%) 6 (14%)
80-89 43 (31%) 20 (21.5%) 21 (50%)
≥90 24 (17%) 9 (10%) 15 (36%)
unknown 1 (1%) 1 (1%) 0
Marital♦
Partnered 65 (47%) 51 (55.5%) 14 (33%)
Unpartnered 43 (31%) 26 (28.5%) 17 (41%)
Unknown 30 (22%) 15 (16%) 11 (26%)
Place of residence§
Farm 24 (17%) 24 (26%) 0
Village (no GP) 10 (7%) 7 (8%) 2 (5%)
Small town (GP, no hospital) 8 (6%) 7 (7.5%) 1 (2.5%)
Town with GP and hospital 91 (66%) 52 (56.5%) 38 (90%)
Unknown 5 (4%) 2 (2%) 1 (2. 2%)
Place of death
Private residence 17 (12%) 17 (18.5%) 0
Local hospital 64 (46%) 54 (59%) 7 (17%)
RACF 44 (32%) 9 (10%) 35 (83%)
Tertiary hospital 8 (6%) 7 (7.5%) 0
Hospice (out of region) 4 (3%) 4 (4%) 0
Other place 1 (1%) 1 (1%) 0
Died in UPoR
Yes 52 (38%) 17 (18.5%) 35 (83%)
No 86 (62%) 75 (81.5%) 7 (17%)
Died within the region
Yes 122 (88.5%) 81 (88%) 41 (98%)
No 14 (10%) 11 (12%) 1 (2%)
Unknown 2 (1.5%) 0 0
*total N includes 4 with incomplete data; RACF= residential aged care facility; ♦ Not possible to determine if
decedent was living with partner; § Not possible to determine RACF place of residence prior to transfer to
RACF; GP=general practitioner; UPoR= usual place of residence; column percentages.
123
When permanent RACF residents were excluded from the data analysis, there were no
statistically significant associations between UPoR and cause of death (p=0.501) or
degree of rurality (p=0.926).
5.6 Limitations
While the data collected are representative of the regional population, there are several
limitations to be acknowledge. Firstly, official death certificates were unavailable, and
while strategies were in place to maximise the data collection, it is possible that an
unknown, though presumably low, number of deaths were missed. While there are
limitations to the accuracy of data obtained through official death certificates (Brameld
et al., 2017), without access to these certificates, verification of cause of death and
interval between onset of the condition directly leading to death, and death itself, was
lacking. It is therefore possible that residents who died of prolonged effects of a
cerebrovascular accident/stroke, were incorrectly excluded. However, the data provided
by the various informants were taken directly from the Medical Certificate Cause of
Death completed by the attending medical officer.
Secondly, because the data collection was regional, the sample size is small. With a
total of 224 deaths, generalisability is questionable. However, the collection of regional
data is both a strength and a weakness. The Australian Bureau of Statistics data (ABS
Regional Stats, 2017c), last updated on 31 March 2017, reported 749 deaths in the
Snowy Monaro, in the five years between 2011 and 2015. Using these data,
approximately 200 deaths would be expected over any given 16-month period. Any
missing records could include deaths of residents that occurred outside the region or
those requiring a coroner’s autopsy, and where no local funeral service was held.
According to the funeral director, only a couple of families withheld consent for the
death details to be released. It is possible that these deaths were still included in the
audit with data obtained from newspaper obituary or funeral notices.
As this study was a component of my PhD thesis, data collection was time limited.
After 16-months of data collection, this component of my research came to an abrupt
halt. Without warning, the main local newspaper ceased publication in May 2016. I had
been gaining vital information from the obituary and funeral notices to supplement, and
124
cross check, the information provided by the funeral director. This reduced my
collection period by 6-months, and therefore reduced the sample size. It is unknown if
the results would be replicated over a different or longer period. While the results
represent where people died in the Snowy Monaro region between February 1, 2015 and
May 31, 2016, they may not be generalisable to all rural regions, or even to this region
over a different time period.
Other limitations include the omission of data relating to income, occupation, and
informal carer support. The results do not indicate decedents’ preferences for PoD,
family caregivers’ satisfaction with PoD, the degree of symptom or carer burden, or the
number of residents that received most of their care at home only to move to hospital at
the very end-of-life. To comply with both HRECs’ requirements, I am unable to add
decedents’ details not obtained through the approved sources. In general terms, some of
the RACFs’ residents moved from their farms months or years before their deaths.
Therefore, their last place of residence was recorded as the RACF, and their farm or
village status changed to ‘town’. It is possible the data collected do not accurately
reflect the previous degree of rurality of some of the RACF residents. While their new
UPoR was the town RACF, many would still consider their farm to be their ‘real’ home.
Despite the limitations, these results describe where residents of the Snowy Monaro
Region died during the same period I conducted the qualitative interviews. These
quantitative data provide a context in which to place the narrative data obtained through
interviews with patients with a terminal illness, and their FCGs and rural health care
professionals.
5.7 Chapter summary
This chapter has described the actual PoD of Snowy Monaro residents, between
February 1, 2015 and May 31, 2016. This was a population survey, with 224 deaths
identified. Nineteen percent of all deaths occurred at a private residence, 29.5% in a
RACF, and 44% in hospital; 85% of deaths occurred within the region. While not all
patients with end-stage cancer, or an advanced non-cancer chronic diagnosis were
necessarily receiving palliative care, 138 Snowy Monaro residents (62% of all identified
deaths), died of a condition considered amenable to palliative care. For this palliative
125
subgroup, 12% died at ‘home’ in their private residence, and 88% died in an institution
(hospital, RACF, or out-of-region hospice).
Highlighting the importance of terminology, when the term ‘UPoR’ was used rather
than ‘private residence’ to depict ‘home’, 38% of the palliative subgroup, including
RACF permanent residents, died in their UPoR (private residence or RACF), with 62%
dying in an institution (hospital, hospice or respite RACF). These results are consistent
with previous studies that report most people die at a place other than their ‘home’ or
UPoR (McNamara & Rosenwax, 2007; Cohen et al., 2015).
Residence, or degree of rurality, was the main factor influencing PoD for the palliative
sub-group, in this region and is consistent with studies included in my ‘Rural place of
death: A systematic review.’ Access to health–system facilities, often lacking in more
rural and remote regions, is a key factor. Cause of death was an influencing factor in
dying in one’s UPoR, but not dying within region. In contrast to the systematic review
findings, age, sex, and marital status were not predictors of PoD. It is possible these
results do not reflect the true picture as some RACF residents may have moved from
their farm more than 3 months prior to their death.
The importance of rural ‘community’ is well reported in the literature. Dying within the
region was possible for 91% of residents. The data do not indicate if dying out-of-region
was by choice or circumstance; however, without a local specialist palliative care
inpatient hospice this may have been the preferred PoD for four residents. Conversely,
the absence of a local inpatient hospice could mean the local rural hospital and RACFs
are considered a substitute for inpatient hospice, and if it is not possible to die at home
then it is important to die within the rural community (Wilson et al., 2009a; Veillette et
al., 2010; Biggs, 2015).
Having determined where terminally ill Snowy Monaro residents die, the findings
illustrate that more residents die in hospital and RACFs than at ‘home’ in their private
residence. What does this mean? The results do not indicate decedents’ preferences for
PoD, congruence between preferred and actual PoD, the reasons behind preferences, the
changing preferences over the course of the patients’ illness, family caregivers’
satisfaction with PoD, the degree of symptom or carer burden, or the number of
126
residents that received most of their care at home only to move to hospital at the very
end-of-life. The following three chapters place these data into context, by exploring the
meaning of ‘place’, the influence of ‘place’ within the concept of the rural good death,
and the influence of rural residency on the ‘good death’ and PoD.
127
CHAPTER SIX: PLACE OF DEATH AND THE
MEANING OF PLACE.
The most important thing is where he's going to be looked after the best.
(Betty, sister of George)
Gwen’s story
As Gwen’s GP, I was the one she consulted when she first developed symptoms.
Preliminary investigations were undertaken locally, and she was found to have a cancer
in her bowel. At her request, Gwen was referred to a Sydney surgeon, as it would be
more convenient to stay with her daughter, both before and after surgery. She was
relieved that chemotherapy was not required. Two years later, she developed liver
metastases, and again consulted the Sydney surgeon. After further surgery, Gwen was
referred to a Sydney oncologist. While it was convenient to be managed in Sydney for
surgery, it quickly became apparent it would not be convenient to travel 500 kms to
Sydney every few weeks for treatment. Chemotherapy was administered in Cooma,
under the guidance of a Canberra oncologist. Gwen tolerated the chemotherapy, but the
cancer progressed. She had chemo therapy four days before my interview. While hoping
for more time, Gwen accepted palliative care with the aim of maintaining her
independence and quality of life.
I met with Gwen at her home on 20th April, 2015. It was a grand, but uncluttered home,
just on the edge of town. The home had been built by her late husband, as had most of
the timber furniture. The house was immaculate. Gwen looked so comfy sitting in her
large leather arm chair. This had obviously been their ‘family’ home and was filled with
photos and other happy memories of her late husband and now adult children. Photos of
grand-children were scattered throughout the home.
It was no surprise that ‘home’ was emphatically her preferred place of care and death. It
was obvious ‘home’ was very important, not just the physical building, but also the
emotional and spiritual tie to family. Her home was comfortable, a place where she was
in control. When first interviewed, Gwen could not “see any problem where I can’t stay
at home.” She already had home help and plans for future support. Family and friends
128
would “rally around.” There would be no need to even consider an alternative, and
definitely not aged care. Gwen was familiar with the local nursing home - it had been
built by her husband. Gwen herself visited regularly to help feed the residents and sit
with those without family, as they died.
Unfortunately, three months after the initial interview, Gwen suddenly lost the use of
her right leg and arm. She was admitted to the local hospital; investigations revealed
brain metastases. She remained positive and while “coming into hospital was not what
we had planned” she “accepted it.” Control of nausea was her main concern. Gwen
now wanted to be in the place best suited to controlling symptoms, while clinging to the
hope of returning home. Radiotherapy in Canberra did not have the hoped-for outcome.
While she experienced small improvements, with no family in Cooma, Gwen was
pragmatic, “I mean I can’t go [home] now and I’m prepared to try whatever happens…I
really want to get home…but I’ll accept whatever happens.” Gwen was transferred to
the residential aged care facility for ongoing rehabilitation. The hope was that given
more time she would gain sufficient mobility to return home, albeit with lots of
community services. She continued to plan her ‘going home’ and what help would be
required. She was conscious of not wanting to be a burden on her family in Sydney, and
friends in Cooma. After three weeks, Gwen’s improvement stalled. She rapidly
deteriorated. Gwen never spoke about going home again.
Gwen had a lovely single room at the nursing home. It looked out onto the garden and
distant hills. The room was adorned with family photos. Her favourite hymns were often
heard on the small CD player. Her family visited from Sydney, often for days at a time.
Initially, between family and friends there was always someone with her during the
daytime. As her body physically deteriorated and she became frailer, the family asked
friends not to visit. During the night Gwen was often restless. A change in rooms was
considered so she would be closer to the nurses. This suggestion caused great distress to
Gwen’s daughter, “I didn’t want her moved…this is her room now…it’s Mum’s home
now.”
On an afternoon in December, while her son was briefly out of the room, Gwen died,
two months after being admitted to the aged care facility. Everyone knew Gwen wanted
to die at home. Her daughter was ambivalent about the impact this had on the quality of
129
her dying. Gwen was comfortable, well cared for, safe and in familiar surroundings, but
not at home. On reflection her daughter said, “I think I was happy at the time with the
decision because she couldn’t have lived [at home] by herself.”
Introduction
Gwen’s story highlights many issues surrounding the actual and preferred place of death
(PoD). Her story demonstrates the issues around disease progression, unexpected events
and scenarios, where remaining in one’s initial preferred PoD is not always possible.
The new place becomes the accepted PoD.
This chapter unpacks some of the issues surrounding place of care (PoC) and PoD, as
narrated to me by the research patients (i.e. those interviewed directly and those
introduced by the three bereaved family caregivers (FCGs)), their FCGs and their rural
health care professionals (HCPs). The chapter begins with a description of the actual
and preferred place of care and death of the research patients. I then explore the
meaning of home, hospital, residential aged care facility (RACF) and hospice in relation
to PoC and PoD, and the congruence between actual and preferred PoD, and between
the preferences of different stakeholders.
6.1 Actual place of death
The previous chapter described where Snowy Monaro residents died. Consistent with
those quantitative findings, of the 15 research patients identified in this qualitative
component of my research study, more died at a place away from their private
residence, (local hospital [five], hospice [one], RACF [five]), than at their private
residence (four). Of the five who died in a RACF, three were permanent residents and
two had been admitted for end-of-life care. None of the patients died in an out of region
hospital. One patient (Gordon) was transferred from the tertiary hospital intensive care
unit (ICU) to die in the local rural hospital.
While these data tell us where the patients died, they do not indicate patient or FCGs’
preferences, nor where they received the majority of their end-of-life care. The data do
not provide any understanding of the meanings attributed to PoD. It was only through
130
the rich data obtained through the interviews that a true understanding of PoD could be
constructed. As will become apparent in the remainder of the chapter, the reality is that
many participants moved between preferences, dictated by the trajectory of their illness.
6.2 Preferred place of death
The interviewed patients, except for two, initially stated their preferred PoD to be their
private residence. One of the exceptions, Barbara, had lived in a RACF for a number of
years, and considered this to be ‘home’ and her preferred place to die:
This is home now…. I’m with people who know me, you know, and it’s quite
happy here. I am quite happy here. (Barbara)
George, who lived in a separate dwelling in his sister’s backyard, stated his PoD would
be either the hospice in Canberra or the local hospital. George highlighted a common
theme amongst the interviewees, and consistent with the work done by Agar et al.
(2008): place of care is not a euphemism for PoD, and often these are two separate, and
different, places. Most patients indicated a short hospital admission would be acceptable
if they required specific symptomatic treatment, such as a blood transfusion, providing
they could return home as quickly as possible. Regardless of where they died, they
wanted to remain home for as long as possible:
I'd like to stay here [at home] as long as I can get myself about a bit and look
after myself. But when it starts to get a bit too hard, on [my sister] especially,
that's when I would go [to hospital or the hospice]. (George)
George’s sister, Betty, did not want him to die at her home. She had nursed her parents
at home and found the burden of care too physically and emotionally draining. So in
George’s case, his personal preference was overshadowed by his sister’s decision not to
care for him at home, and his willingness to respect her wishes:
I don't know, it sounds horrible, but I just can't do it again. I just can't nurse
someone to the end again, it's just too much. (Betty, sister of George)
131
Two of the three non-participating patients (i.e. their FCGs participated only in a
bereavement interview), had indicated a preference for PoD to their FCG. The third
patient (Gordon, who I will discuss later) had become unwell suddenly, without
discussing end-of-life goals with the family. According to one FCG (Andrew), his wife
wished to die in the out-of-region hospice because:
She didn’t have confidence in me to be able to manage the stuff right … Elaine
was a nurse, and I’d work out the medications and everything my way, and I had
tables drawn up and the whole thing, but as she got sicker she got more
confused and less confident that I was doing the right thing. She would have
been happy if there was somebody looking over my shoulder that could say, “Oh
yeah, he’s doing it right.” (Andrew, husband of Elaine)
Another patient, Leanne, had told her partner that she wanted to die in the local hospital
rather than on the farm. It is uncertain if this decision was guided by her concern for her
partner, or reflected her own wishes:
We talked about things as things went along and…I’d made the decision that I
didn’t want her to die in the house. And I think Leanne probably was thinking
along the same lines as well…I didn’t want that lingering memory flashbacks in
my head of her lying dead in the bed. I mean I can still picture her in the
hospital and I’m just glad that that memory’s not in [the bedroom]. (Daniel)
However, when the time came to leave the house, Daniel struggled to phone the
ambulance, for he knew Leanne would not return home. Daniel recalled the
conversation with his partner before making the call:
Leanne had to tell me three times to ring the ambulance. She said, “Have you
made that call yet?” (Laughs) and I said, “No”. So then the third time she said,
“Get the phone and make it and do it here where I can hear” So yeah that was I
guess just a realisation that things were not so good. And I know when she did
go in [to hospital, the GP] said if you feel you get a little bit stronger … you can
go home for a day or so. And she just said, “No I’m not going back there, I’m
132
not going through that again.” So obviously, even though she put on a brave
face when she was leaving it was fairly tough on her. (Daniel)
Drawing out the preferences from the participants was not a straightforward matter and
highlights the limitations of only asking a single question in a survey or questionnaire.
While some patients gave an emphatic initial answer in stating their preferences, usually
“home”, most qualified their response during the interview. While hoping to remain
home, most were prepared to wait and see how circumstances unfolded:
Depends on how bad the pain was or his hallucinations and things like that. If
he was to go into a coma he would probably be better off in hospital and I’d stay
at the hospital. I mean hospital’s great, but only if he needs it. And if he was in
really bad pain, yes. (Carol, wife of Ryan)
Amongst this rural group, the preferred PoD did not only refer to the physical building.
Except for Elaine, who chose to die in the Canberra hospice, some 115 kms from home,
all participants wanted to die within the local community. George had moved from
Canberra five years previously, so the option of dying in the Canberra hospice was not
totally ‘out of region’. For some, dying in the local hospital softened the disappointment
of not dying at home. While Ivan did die at home, he summed up the importance of
remaining in the community:
Well I'm part of [the community]. So why would I put part of myself somewhere
else? You are what you are because you make your life where you are. (Ivan)
Gordon’s family arranged a transfer, for end-of-life care, from the intensive care unit
(ICU) in the tertiary hospital to the local rural hospital. The family wanted him back
home in town. Home as a physical building was not the focus. ICU was a distressing
place for the family. Gordon had not regained consciousness and was dying. The family
wanted their father out of ICU, away from the tertiary hospital, and back into the
familiar surroundings of the local hospital, “where he could die peacefully” and they
would be supported by familiar hospital staff.
133
Preferred PoD was often a place where patients felt valued. Barbara was a long term
resident in one of the region’s RACFs. Visits from family were limited, as none of her
children lived nearby. She agreed to relocate to a Sydney RACF, so at least one son
could see her more frequently. However, leaving her community was not a good
decision and she became “homesick for [the nursing home] and that’s why she went
back [to Cooma], she wanted to go back…so in the end it was really where mum was
happiest,” (Rodney, son of Barbara).
While Gwen had earlier travelled to Sydney for surgery, so as to be with her daughter,
placement in a Sydney RACF was never an option. Louise knew that if Gwen could not
die at home then it was important she remain in her community, surrounded by friends:
I would never have taken her to Sydney, she’s got all her friends here, and she’s
familiar with here and knows everyone. I know how hard it was for her when she
was in Canberra for those 5 days, she really missed having people she was
familiar with looking after her, so I would never have moved her to Sydney…If
I’d had to move her to Sydney that would have been awful for her, I think that
would have broken her heart actually. (Louise, daughter of Gwen)
This desire to die in community was not universal. Elaine chose to leave her community
in order to die in the place she felt was most equipped to manage her symptoms:
Her choice was [the Canberra hospice], after having such a good experience
[previously]...The pain control team there is very, very good. (Andrew
discussing wife’s preferences)
Each participant had a preference for where they, or their family member, lived out their
last days, and where they died. They all had legitimate reasons for their preferences.
The following sections of this chapter explore the meaning behind each place identified
by the participants as a place to die.
134
6.3 The meaning of home
Most patients, including the permanent RACF residents, stated ‘home’ to be their
preferred place of care and place to die. While patients and FCGs identified home as the
physical building, the words and gestures suggested home was more than that and was a
place of security, memories, identity and control. Many of the interviews were
conducted at people’s homes, and just visiting where people live gave me an
appreciation of what ‘home’ meant. Often I could hear the birds singing, the sheep
bleating in the paddock, the neighbour mowing the grass or the sound of silence.
Hearing his son on the tractor gave Ivan a sense of connection to the everyday running
of the farm:
Ivan: I like lying here in this bed. I’ve got a lovely view out the window, and the
garden. I've got my cat. And I got my two dogs with me. And my chooks there.
So it's a perfect set up. And then I can ring Paul [son] up and say, what have
you been doing Paul [on the tractor today]?
Interviewer: So even though you're in bed you're still part of the activities on the
farm?
Ivan: Very much part of it.
For Ivan and Bruce, home was the generational family property where they and their
FCGs felt a connection to the land and a continuity with the past:
It’s sort of about who we are really, which might sound silly, but a city person
wouldn’t have that connection with place that we have. We’ve been here six
generations now, so when we look at things around the farm, like if I look out
there I can see a pine tree, I know my great grandfather planted that. I go to the
post office and the holly trees out the back, my great grandmother planted
those…. Yeah, so we’ve got connections all over the place. (Paul, son of Ivan)
For others, home was also the place for keeping important memories, “if anything
important was done it was always done here [on the farm]” (Bruce). After regular
home visits, spanning more than two years, the community nurse observed that
135
Marianne’s identity as a mother was embedded in the home. The farmhouse was full of
memories of children growing up:
That property was where she’d done her mothering, where she’d educated her
children, because she didn’t send them to school, so for her that was her world
really, so for her dying in that setting was I would imagine very important to
her. (Marianne’s community nurse)
At times the words used to describe ‘home’ were paradoxical. On one hand ‘home’ was
described as peaceful, unregimented and yet it was a place where patients, like
Marianne, felt in control:
[Home is] a controlled space that Peter [husband] and I have control of, where
as soon as we go into a hospital we’re in somebody else’s controlled spaced.
That’s what immediately springs to mind. (Marianne)
Proximity to town contributed to the ‘controlled space’ and the security of home. While
living on a farm had disadvantages, as will be discussed in chapter 8, Peter appreciated
the isolation that the farm provided.
This is a good location because it’s a nice place to be, if you’ve got to be sick,
you know, you’re not in suburbia, stuck in a street with people calling in all the
time. I mean, you can manage the visitors [at home]. (Peter, husband of
Marianne)
For most patients, quality of life was maintained if they could compensate for the loss
of control over their disease, and its effect on their bodies, with some degree of stability
and ‘normality’ in other aspects of their life. Many participants, like Ray and Carol,
spoke of being able to continue normal activities at home:
Well probably being in surroundings that you know, not having to depend on
too many people to do things, like going to the toilet and things like that, just
little things. (Ray)
136
We spend a lot of time at night just talking in bed. Just cuddling and talking.
We’ve always done that anyhow but more intense now. (Carol, wife of Ryan)
The close connection to a ‘place of memories’ facilitated the representation of home as
a place of independence and individuality, two qualities that Gwen fiercely protected
before her physical deterioration forced her to transfer to a RACF:
Home means everything to me… I’m a homely person, I’ve always been home
and family, and it is more personal, its more… it’s more intimate and I think that
you can, when I say relax I don’t mean relax, but you can feel more comfortable
at home, you can do what you feel you should do. (Gwen)
Home was not simply the physical building, and did not refer to a ‘private residence’ for
all the patients. Three of the patients were permanent residents of a RACF; two were
short term residents. The presence or memories of family did not necessarily define
‘home’. For Barbara, who had left and then returned to the local RACF, the facility was
her physical, emotional, and spiritual ‘home’, a place where she felt she was treated as
an individual. Her family did not visit regularly, and yet when she left Cooma to be
more accessible to her son, she pined for her former ‘home’:
At first she was doing very well and we thought [we’ve] found the right
place…Big relief… but we noticed her health started to deteriorate and it wasn’t
so much her physical health, … it was getting her down, and she said, “I want to
get out of here.” (Rodney, son of Barbara)
So after “a couple of weeks … I thought, no I don’t have to put up with this kind
of nonsense, I’ll go back home. I call this [Cooma RACF] home… this is home
to me.” (Barbara)
Morris and Thomas (2005) observed that while home is a familiar place, where patients
and their FCGs strive to maintain a sense of normality, it becomes ‘less normal’ as the
familiar space is altered to accommodate the various supports required to keep the
patient at home. Kevin’s community nurse observed that Kevin and his family
accommodated the ‘new normal’:
137
Because Kevin had been so sick for such a length of time it was normality for
them… to have him… at home. I know it sounds a bit bizarre. But it was normal,
it was part of their normal life, he’d been like that for so long (laughs). The
progression to end stage was very normal, not much had changed. He might not
be responding as much, but the physicality of having him in a bed and caring for
him in a bed, normal. And maybe that’s why it [went so well] … it was what he
wanted and that’s what Deedee said to me. She said “thank you, that’s what
Kevin wanted and thank you for being able to help us with Kevin and provide his
wish.” So she was certainly very grateful of that. (Kevin’s community nurse)
Home was often idealised and portrayed as an unproblematic place, and while the
primary preference for most, home was also offered as a default from unacceptable
alternatives or places considered problematic. For Gwen, institutions were viewed as
impersonal; for Kevin’s daughter, hospitals were too noisy; Peter was concerned about
the loss of control in a hospital setting and the staff’s inability to manage his wife’s
medications:
I have thought about [the alternative to home] but it does not appeal to me in
any way, because it’s not personal. If you go to a nursing home, or where ever
you go to, it’s not personal...If you’re in an institution you’ve got to abide by the
rules of that institution. (Gwen)
[Dad] likes quiet, peace and solitude, so I think a hospital where people are in
and out and it’s busy, [means] he wants to be [at home]. (Susan, Kevin’s
daughter)
There are so many things about hospital care that don’t suite this disease state,
unless she was completely unconscious, and then it wouldn’t matter, but whilst
she was conscious she would suffer terribly with the anxiety and worry about
“will my medication be on time, will I have someone to do this and do that?”
We’d be at the point of having a family member sitting with her in hospital when
we could be doing it at home. (Peter, Marianne’s husband)
138
However, home is not without problems. Morris and Thomas (2005) observed that
normality, security and personal control found within the familiar environment of home
can quickly become disrupted. When symptoms escalate and access to medical help is
not readily available, especially after hours, or the burden of care becomes too great, the
safe home can quickly become an isolated place, where patients and FCGs feel
abandoned, and FCGs become overwhelmed with the responsibility of caring. Even
when things seem to be working well, caring for someone at home is exhausting for the
carers. In the following quote, Ivan’s son describes the exhaustion faced by his sisters in
providing the physical care for their father. He also illustrates the challenges faced by
FCGs in making medical decisions without a clear understanding of the medical details:
There is stress on family relations, everybody’s under pressure, exhausted. It is
exhausting, there’s no let up. Dad would be up six times a night… I eventually
worked out a roster. One [sister] would do night shift, one would do day shift.
They were doing eight hour shifts toward the end, to try and rest, so it’s
physically exhausting from that aspect. Another difficulty is managing all the
information, so who’s right and who’s wrong, medicine’s not a perfect science –
should we be doing this? Should we be doing that? What’s got to happen here?
What’s got to happen there? And managing that is difficult…One of the biggest
issues was just understanding what’s going on, the information from
[everywhere], trying to process it without the background and being able to
speak the language, let alone understand. (Paul, son of Ivan)
Symptom management is often more difficult at home, especially in rural areas where
medical and nursing backup is often not available, especially after hours. While
Marianne settled in the last few hours before her death, her GP considered hospital to be
the only option had her terminal agitation persisted:
I think the fact that they didn’t have medical staff there to provide reassurance
and intervention as necessary was certainly challenging… She became more
agitated over the next few hours and there was no medical person there to
intervene and help them with that…. If we hadn’t been able to [make her
comfortable], she would have had to have come into town. I mean you couldn’t
have her distressed in her last hours. That would have been awful. I mean that
139
would be a very distressing thing… you couldn’t possibly have kept her at home
in the discomfort and the distress. No that would have been cruel to do that to
her and to the family. (Marianne’s GP)
Ryan and Carol were both keen for Ryan to be cared for and to die at home. From past
experience, through caring for a relative, they had the image of Ryan slowly slipping
away, peacefully at home with friends and family all around. The tranquillity of this
image unravelled when Ryan became acutely confused and agitated.5 The palliative care
nurse describes a scene of chaos and distress when she arrived at their home on the
morning Ryan died:
Things were going well leading up until his death. They…had seemed to
manage well, and I think his wife was quite controlled. But the morning of his
death, when he became agitated, extremely breathless, the oxygen that we had in
place wasn’t enough, the pain, he had pain, extreme pain. I think everything...
all his symptoms exacerbated at about early morning. I think they had been like
that for quite some time, probably maybe from about two a.m. until we got there
…about ten o’clock…We had anticipated ...so the drugs were all there…. I think
[Carol] was quite frightened at his agitation, so if we’d both not been able to be
there and start the drugs I think she would have been relieved to have gone back
to hospital, because she would have been... relieved to see him more
comfortable, wherever that took place.
I spoke to [Carol] on the phone the next morning, and I said, “I’m so sorry that
Ryan died on the floor.” But, I was very happy when we got [Ryan] back into
bed because he looked really peaceful and [Carol] was then able to lay with
him. And [Carol] said, “I was just so happy to be able to lay there all day with
him, and that is what I’m remembering.” So that made me feel better. (Palliative
Care Nurse/Ryan)
5 This was one occasion that I was present at the death in my role as the palliative care doctor. I also
lived down the road, and as luck would have it I was home when Ryan’s sister phoned me to visit
urgently.
140
Most participants spoke of the importance of continuing social engagement
with extended family and friends. This was often considered a vital support for
the patient and the FCGs. In the following quote, Carol expressed the important
role played by maintaining social contact at home, and the challenges of
maintaining this support if Ryan was hospitalised:
When you’re home you can have a house full of people and they can come and
go at different times. Not all the time, but then you’ve got the support with you
the whole time. Where in a hospital you haven’t…and you’re more aware of
making noise and disturbing other patients. Where at home, you know I can
have family in their caravans come and stay in spare rooms, put mattresses on
the floor. You could have the house full. Not necessarily disturbing his sleep, but
they can pop in from time to time, but they’re also there for me when I need a bit
of a break as well…you just can’t do that in a hospital. (Carol, wife of Ryan)
Not all participants wanted their home to be “full of people.” A few months before
Marianne died she became gravely unwell. At the time, all the family were called home.
Their home was full of adult children, their partners and Marianne’s grandchildren.
With no one with medical or nursing experience, the home became chaotic. Marianne
did improve, and life returned to normal for another few months. On reflection,
Marianne’s husband said:
The home can’t become the social hub, it has to become a private hospital where
there is only one family member and they’re here to care for her, not to talk to
others, or to socialise with other family members or catch up or anything like
that so that Marianne only has to cope with one person coming and going and
looking after her…if I’m not here. (Peter, husband of Marianne)
While most participants identified home as their preferred PoD, the decision was
complex, dynamic, and often required much negotiation between the patient and their
FCG. In many instances, the role of the HCP was to help navigate through the options.
The meaning of home, as described by the participants, was consistent with the ten
features identified in Despres’ (1991) literature review, and is more than just the
material structure. For the research participants, home was a place of security and
141
control; a reflection of one’s self, value and status; a place of permanency, continuity,
and a place to belong; a place of memories; relationships; the centre of activity; and a
refuge from the outside world. All 15 research patients remained home for some time
during their illness, yet only four died in their private residence. The following sections
of this chapter describe the alternatives to home.
6.4 Hospital as the place of death
Two patients were interviewed in hospital, with the remaining patients providing
reflections on their previous admissions, or on their hopes and fears for possible future
admissions. Their experiences included the local rural hospitals, as well as tertiary
hospitals in Canberra and Sydney. Five of the research patients (George, Leanne, Ray,
Stuart and Gordon) died in one of the two rural hospitals. Their FCGs discussed, in the
bereavement interviews, their experiences of hospital end-of-life care. While mostly
positive, there were some significant challenges and disappointments; however, all said
the hospital was the right place for their family member to die, and if faced with a
similar situation they would again chose the hospital as their preferred PoD.
Patients were transferred to hospital for two main reasons: (1) symptom management
and (2) to relieve the carer of burden and responsibility. This second reason was either
pre-planned, as in the case of George and Leanne, or became necessary during the
illness trajectory. For a home death to be successful, symptoms need to be well
managed, and the FCGs need to be confident in their ability to care:
I thought for Kevin he was certainly very comfortable… he had a very
supportive family that worked really well as a team. His main care provider,
Deedee his wife, was very happy in that role. She was very good at keeping
accurate records (laughs) which made it easier for us [nurses]…So yes I felt that
[home] was the right place for Kevin to be, yeah. And I felt the family and the
carers were very, very happy with Kevin being there too, and in their role as
well. (Community nurse)
While hospital was not the preferred PoD for the four patients who died at home, during
the interview process they all indicated that the local hospital was a genuine alternative
142
if symptoms could not be managed at home, or if their carer was not coping. Ivan had
decided to leave that decision “entirely to my doctor.” One GP, speaking in general
terms, felt that too much emphasis was placed on achieving a home death and that often
hospital was the more appropriate place:
I think that [patients] don’t realise the stress that it can put on their carers and I
think there’s probably times they can actually suffer more than if they were in a
hospital environment where the professionals can respond to the rapid changes
in their need for pain relief in particular, which you just can’t offer at home on a
24/7 basis. (GP)
My first interview with Ray was during an emergency hospital admission due to
increasing breathlessness. Ray was hoping to return home, but after two failed
discharges, home was looking more unlikely. Ray was aware of the stress on his wife,
and eventually her wellbeing became more important than getting home. Ray did not
return home, and in the second interview he told me was now “more sensible about not
going home…as there was no way Clare could care for me now.” Clare was relieved in
his acceptance of the hospital:
I panic… I’ve had enough. He’s had enough and I’ve had enough, so the nurses
and the doctors can look after him here…I get very tired….The hospital can do
more for him than I could at home because I… you know, I haven’t got the
oxygen that he needs…he’s better off here. (Clare, wife of Ray)
While it is acknowledged that rural hospitals lack the palliative care resources and
expertise of the larger city hospitals (Lynch, 2012), the Snowy Monaro hospitals have
their own standout features that partially make up for any deficiencies. Patients and their
FCGs spoke of the personal treatment by all the staff. The hospital was a familiar place
where some had worked and knew “most of the staff” (Nathan, son of Gordon), a place
where children and grandchildren had been born, family had been cared for, and former
family members had died, the place they turned to whenever a crisis arose.
143
One GP recounted a recent interesting experience:
I had a patient die in hospital…his wife was actually in hospital at the time with
a problem with her hip, so she was up one end of the ward and then they called
her to his room, and she was with him when he died, which was nice.
The following quotes by Ryan, Clare and Stuart illustrate their experiences and
perceptions of rural hospitals. Ryan had recently spent a few days in Cooma Hospital,
for symptom management; Ray and Stuart were local hospital inpatients at the time of
interview:
I prefer [the local hospital], because I just like… country hospitals…and
country people…. I feel they’re more compassionate. They’re not as highly
staffed so there’s more personal interaction. In the big hospitals there’s that
many nurses and doctors that you don’t get to know them or their name or
personality… there’s a definite difference between the country hospital and the
city hospitals. (Ryan)
You get along better in the smaller community hospital rather than busy, busy
[city] hospitals. The hospital has been very good to us. I don’t think he’d get any
better care than what the hospital’s doing for him here. (Clare, wife of Ray)
It's much better [here]. Canberra's really busy and because it's a teaching
hospital there's people coming and going all the time. Everybody wants to have
a look. (Stuart)
While most participants highly praised the nursing staff, all the staff contributed to
making the hospital more personal. Stuart’s brother, Fred, had come from Sydney to
care for Stuart. He was impressed by the country hospitality:
All the staff, the tea ladies, the boys that bring lunch and everyone, the whole
thing seems to be done in a very respectful way… and then when you meet them
out in the grounds having their lunch or something, they chat. (Fred, brother of
Stuart)
144
Ryan even laughed about the advantages of knowing the kitchen staff:
I know [the assistant] in the kitchen quite well. I’ve known her for sixteen
years…she’s a partner of one of my mates. She come in and ask me is there
anything special (laughs) I wanted on the diet and so ordered some scrambled
eggs for a change, and all of a sudden I’d get scrambled eggs whenever I
wanted them. Yeah, little things like that…you wouldn’t get that in a big
hospital. (Ryan)
This notion of ‘knowing’ and ‘being known’ (Pesut et al., 2011) by the staff was an
overriding expectation of all the participants. Leanne and Daniel had a great rapport
with Leanne’s GP and community nurse, both of whom visited her on the farm, 30 kms
from town. What occurred after Leanne died in hospital was unexpected, especially in a
small rural hospital, and caused great disappointment amongst the family. The doctor
involved was the overnight medical officer, a GP rostered to work in the emergency
department and ward. He was unknown to Daniel and had not been involved in
Leanne’s care:
Just one thing I was a bit disappointed about. One of the nurses was in there just
after Leanne died. She just came in and checked her and she said “I’ll have to
get the doctor to certify everything”. So we waited there for probably 20
minutes… and then the doctor come to the door but he didn’t actually step into
the room, he just went to the door and he asked was the nurse in there. We said
“no” and he went back to the desk, and then we waited another 20 minutes, and
then we went out and said to the nurse, “You know the doctor hasn’t been.”
“Oh yeah he’s already signed off”, but he didn’t go into the room…which was a
bit disappointing…I would have thought he would have, might have offered
some condolences or something. He didn’t even look at Leanne. (Daniel)
The familiarity with staff did, on occasions, have its downside:
I love the oncology unit… you know people there, you go, you chat,
you talk it’s really friendly and lovely and I used to love going but [the
nurse] would [speak] at the top of her voice. It wasn’t a private thing
145
and I didn’t feel comfortable at all… Other patients, we can chat and
talk and staff shouldn’t say anything. They are actually there to do a
job. (Marianne)
With only a small number of staff, when conflicts arose, there was often no alternative
health provider to turn to. This conflict with a staff member eventually caused Marianne
to cease local treatment, “I couldn’t take it anymore” opting to continue treatment in
Canberra. Other participants also had negative experiences in the local hospital, often
surrounding the prescribing and administration of medications. Andrew spoke of the
inexperience of medical staff in dealing with high dose opiate pain medication; Susan
spoke of the delays experienced when her father required medication:
But they wouldn’t [give her what she needed] because they’re not trained in [the
medication] or the methodology. (Andrew, husband of Elaine)
In hospital you can’t give the medication as the carer, you can’t be in charge of
their pain… let’s say [Dad] might need the lolly [fentanyl lozenge], but you
might ring the buzzer, wait ten minutes, then they have to go and unpack the
lolly and even then Dad was showing them how to unpack the lolly. I think it’s
quicker at home because within three seconds of Dad saying, “Can I have
something?”[he gets it], and he knows how to open the lolly better than the
nurses in the hospital. (Susan, daughter of Kevin)
However, this negative experience was not universal across all patients as
expressed by Ryan:
If I needed anything for pain… Not a problem…They were always there, yeah.
Excellent service. (Ryan)
The delay in receiving medication was also indicative of the low staffing levels, not
unique to rural hospitals, ‘it’s the same everywhere” (Ray). While the level of care was
perceived as mostly good, families sometimes felt the care lacked expertise compared to
city hospitals, where care seemed “more proactive rather than reactive” (Daniel,
partner of Leanne). Some families also felt the need to remain with their family
146
member, even overnight, to ensure they were adequately cared for and received timely
assistance:
I guess [the care] was good.…they’re all busy…like you know Leanne’s
breakfast had arrived but she wasn’t able to open anything or eat it so that was
a bit of a wakeup call to the rest of us… then we just said, well someone should
go and sleep there every night and be there for whatever she needs. (Daniel,
partner of Leanne)
This need to stay with their family member was for some an inconvenience and one
reason some FCGs preferred to continue care at home:
But I really think that [a hospital admission] would just be more hassle because
we’d still want to be there, one of us there all the time or two of us there all the
time, so that’s actually much less convenient. (Susan, daughter of Kevin)
With no formally designated palliative care beds, Cooma hospital does have two rooms
set aside for palliative care patients, with facilities for families to stay overnight in the
Mary Green Room. Despite some families perceiving the overnight stay in hospital an
inconvenient necessity, for some families, having the ability to stay overnight was a
privilege and lessened the disappointment of their family member not receiving end-of-
life care at home. One GP describe his experience with a family who were able to be
with their husband/father when he died in hospital:
His wife said she couldn’t really cope and, we’ve got this room upstairs that was
furnished by the MCCR [Monaro Committee for Cancer Research] and the
family were there, she was with him, and the kids lived [out of town] so they
were actually in the hospital and it was a very good hospital death… they,
physically, couldn’t have done it at home. (GP)
One theme, throughout the interviews, was the importance of maintaining
personal routines and normality, regardless of place of care. While hospitals
were considered a ‘controlled space’ where routines and policies were adhered
to, the hospital policies often were not sympathetic to individual preferences.
147
As a controlled space, and in consideration for other participants, some felt the
hospital was not a congenial place to maintain social interactions. Ray was
concerned about his noisy grandchildren visiting:
All the grand kids can come into the house. If they come here [hospital]
(laughing) they’ll wreck the place, they’re little ‘ferals’! (Ray)
The lack of privacy in hospitals was perceived to inhibit intimacy between
couples. For Ryan and Carol, it was important they have “that special time…
[when] we’re not interrupted… that quiet time, that personal time.” Ryan died
at home and the time Carol spent with Ryan before the funeral directors came
was important, as narrated by the palliative care nurse:
And for [Carol] to be able to spend quite significant hours, because [Ryan]
stayed in the home until about seven o’clock...until the daughter arrived, so she
was able to lay with him. They were in their own bed. [In hospital] it would
have been possible for a little while, but it’s still a foreign environment, a
hospital environment…And people coming in and out of the room, or
noise…Yeah, just hearing that background hospital noise, even with your door
shut. And I think that after a little while [the nurses] would have been
suggesting that Ryan go down to the morgue. (Palliative care nurse/Ryan)
Despite the challenges and disappointments with the local hospitals, the FCGs
of the five patients believed the hospital was the right place of death:
Hospital was the best place...yeah I think so. (Daniel, partner of Leanne)
Even though he wasn’t at home. I thought it was good actually because I don’t
think we could have coped at home…So, I think [Stuart] did it for us. He
probably thought right, I know them up at the hospital, I know everyone there …
I’d rather die there where they can look after me than expect my brother and
sister to do it for me. (Fred, brother of Stuart)
148
After a month long stay in intensive care in Canberra, Gordon was transferred to the
rural hospital to die. While returning home was considered, the better option was a
transfer to hospital. His death in hospital was managed professionally, compassionately
and personally:
He was given a private room where he could die peacefully, with his family
around him and I think that was as dignified as it could be. He had people
tending to him… he was off the drugs, but I think he was just getting oxygen...if
we wanted to have him home in our house then we probably could have, but
we’d made a decision that he was just as well off being there [in hospital] as
being at home…it would have been harder having him at home, because we
would have had to do all the work…it was fine in Cooma [Hospital], so why not
stay. (Nathan, son of Gordon)
In contrast to the well managed hospital death of Gordon, all did not end well for
George. The reader will recall that in respect of his sister’s wishes, George’s preferred
PoD was hospital. George had had a number of admissions to hospital, the last
admission the result of a delirium. All through his illness, George had known he was to
die in hospital; however, in his confused state he wanted to go home. He became
agitated and at times combative, especially after his sister’s visits. Possibly due to the
inexperience of rural staff and the lack of specialist palliative care advice, the staff at the
small rural hospital requested that Betty stop visiting. I will let Betty tell the rest of the
story:
George wasn’t really happy in hospital. He wanted to come home… [When I
visited he’d say] “I want to go home” and I’d say, “Well, you can’t come
home.” “Why can’t I come home?” “Because they’re not going to release you
and ‘cause you’ve got cancer and you’re dying and you need to stay here.” They
stopped us from seeing him, from the end of April to, he died on the 9th of May.
We didn’t see him in that time...’cause I used to upset him every time he’d see
me. He’d get upset because I wouldn’t take him home.
At ten o’clock I got a phone call from the nurse, who told me that she walked
into George’s room and he was a bit blue, so they put him in bed and they put
149
him on a syringe driver and she said, “Now, that doesn’t mean he’s going to die
today. It could be three weeks’ time or whatever.” Twelve o’clock I got a phone
call telling me that he’d passed away… I mean, I wasn’t there, I didn’t see him,
so…obviously, his condition had changed, so the fact something had changed to
prompt them to change his management. I mean, what she said is that sometimes
people will change and will go on for a few days, so she’s right and she’s
wrong, but yeah, she was wrong.
He passed away on his own, which he didn’t want to do. We’d already
prearranged that when his time was up or getting close to being up, I would be
there with him because he said he didn’t want to die on his own. “I will be there
for you. I will hold your hand until you pass away” and he said, “You
promise?” and I said, “I promise.” But I didn’t get the chance to do that.
(Betty)
When I interviewed Betty, three months after George died, she was finally coming to
terms with the guilt and trauma of not being with her brother when he died. She knew
she could not have cared for George at home, but she was disappointed she had been
stopped from visiting. Betty was left wondering if things would have ended differently
if George had been transferred to the out-of-region hospice.
Not all research patients admitted to the local hospitals for symptom management or
carer distress remained there for end-of-life care. When their prognosis was estimated to
be more than a couple of weeks, and they were unable to return home, then the only
option for Gwen and Bruce, was transfer to one of the six local RACFs.
6.5 Residential aged care as the place of death
Five participants died in a residential aged care facility (RACF). Three were permanent
residents, for whom the facility had become their ‘usual place of residence’. Gwen was
admitted for respite and rehabilitation, and Bruce was transferred from hospital for end-
of-life care. For Barbara, introduced previously, the RACF had become more than just
the physical building, it was also her ‘home’. While Cheryl wished she could be at
home rather than in the aged care facility, she “sort of” accepted the placement and had
150
made it more homely by bringing in “some furniture and stuff.” Dorothy, on the other
hand, struggled with her placement; the facility was not her true home. All of these
participants had moved to the nursing home, not from choice, but because their health
and/or social circumstances meant they were no longer able to be cared for at home:
She never wanted to be put in a nursing home…but unfortunately she needed 24
hour care and it just wasn’t something I could do. (Rhonda, daughter of
Dorothy)
Two RACF research patients were widowed (Gwen and Barbara); three were married
(Cheryl, Bruce and Dorothy). While Cheryl’s husband was willing to care for her at
home, her condition required round the clock nursing care, not available in her out-of-
town residence:
If she could move she could walk, she wouldn’t be here and could leave, that’s
the unfortunate part about it, she could be at home….but, we wouldn’t have the
equipment for a start, like she’s got to be lifted in and out of the chair all the
time with the lifter and that takes virtually two people, so you would have to
have someone with you all the time. (Pat, husband of Cheryl)
The remaining residents had either no partner or the partner had their own health issues
making it impossible to provide care at home:
Well there is Dad and he’s 91 and has mild dementia. Up until Mum went to
hospital I was coming over all day and being with them all day. But …I got sick
so I couldn’t keep on doing it…because of her behaviour it was becoming really
unsafe for her [at home]. (Rhonda, daughter of Dorothy)
Families often struggled with the decision to admit their family member to care.
Hospital was a more acceptable place than a RACF. The patients’ medical condition
was seen as the reason for admission to hospital, the patient required medical care that
was not available at home; the family’s withdrawal of care, desertion or shirking
responsibility was the perceived reason for admission to a RACF, and this was often
associated with a sense of abandonment and guilt:
151
I felt it hard because I thought I was letting [mum] down, see, after all the time
looking after her that I actually let her down by putting her in a nursing
home…When I told her that’s what we had to do, she said “I know it’s not your
fault”, … but I don’t know whether she fully understood. (Rhonda, daughter of
Dorothy)
According to one RACF director of nursing (DON), residents take on average 3-months
to settle into their new environment, and Dorothy had only been there two weeks. Bruce
was transferred to the RACF from the local hospital. He had a mild dementia that
worsened as his cancer progressed, making it impossible, and unsafe for him to return
home to his elderly wife. During his 2 month stay he never settled, the facility never
became home, his family never resolved their guilt. In contrast, Barbara had lived in
aged care for a number of years. According to Barbara and her son, Rodney, there are a
number of features that make a RACF stand out. Number one is the staff:
Some of the staff there [in Sydney], you could tell it was a job and their attitude
was quite different, so I think the number one thing that makes a nursing home
better, is… the staff, it’s like the foundation of it. (Barbara)
Décor and the airiness and all that, and the layout, come in second. I think the
big thing with [the Cooma nursing home] is its staff are really dedicated. I
can’t think of one person there that I’d say was not doing their job. (Rodney,
son of Barbara)
Staff aside, if the RACF is to become home, then it is important to make it less
institutionalised and more homely. Rodney and Fred described their experiences of
nursing homes in Sydney:
It was dull, it was depressing… I just thought, the designer has missed the mark.
The painting, the paint inside, the walls and décor, was all that dark coffee
colour. (Rodney, son of Barbara)
When I go into nursing homes, when I used to visit Mum, the nursing home just
smelled. And it had a kind of depressive side to it… and everything associated
152
with death…On the whole I think probably rural nursing geriatric care in some
ways is better than in the city. (Fred, brother of Stuart)
One theme that was often voiced, was that country RACFs are staffed by community
members, people who know the residents and whom the residents know. For many new
residents it is a familiar face that helps them to settle in:
The staff really care, because they know the residents. That’s the difference
between rural and city, is that the staff themselves have grown up and lived with
them. They know them, so they do the death [the way they] think [the resident]
would want. (DON)
From my experience with nursing homes in Sydney, I think there’s much more
community here and the staff knew mum and… in a way it can be really
supportive that they kind of knew her. (Louise, daughter of Gwen)
Just like rural hospitals, rural RACFs are part of the community. This can work for or
against its reputation. If something is done well then the facility is held in high regard; if
there are gaps in care then the whole community knows. The family members who were
interviewed commented that at times “they were really quite short-staffed” (Louise,
daughter of Gwen), and in particular short of qualified nursing staff. This became more
of an issue when the resident was unable to call for help themselves and relied on FCGs
to assist:
We wait a long time when the buzzer’s pushed for help, for the pan or for pain
medication. I think the staff probably do as best as they can but there’s times
when I’m very frustrated by how long it takes to have someone come. (Rhonda,
daughter of Dorothy)
I don’t think the nurses read up so much ‘cause a lot of them don’t understand
mum’s got macular [degeneration] … they tend to shout at her and I think well
she’s not deaf, she’s blind. Mum does get a bit of a pip in her when they start.
(Simone, daughter of Dorothy)
153
However, Cheryl’s husband did not see that staff shortages were an issue at his wife’s
facility. Compared to their experiences in a city hospital, Cheryl was well looked after.
Pat felt very confident in the level of nursing care:
I can go home at night and sleep knowing she’s well looked after. When she was
in Sydney I’d be worried every 5 minutes wondering if she is being looked after.
So much turnover of staff in those hospitals up there, not regular, they come
from the agency and you never know who, never see the same one twice. (Pat,
husband of Cheryl)
It was only possible to interview two FCGs in the bereavement period; however, both
felt that the facility was the right place to die. All the residents died in a safe
environment. They were all well cared for. They died in a RACF because home was no
longer safe for them:
It was the right place, she was surrounded by friends, but also not only staff and
patients but there was other friends within the town who used to come and see
her…so she felt good about that. (Rodney, son of Dorothy)
I was happy at the time with the decision because she couldn’t have lived [at
home] by herself, and at the time we didn’t know how long she had. (Louise,
daughter of Gwen)
While FCGs would consider a transfer out of the facility to hospital “if they
could do something” (Louise, daughter of Gwen), none of the residents or their
FCGs wanted to be transferred out of the facility for end-of-life care:
I wouldn’t want her moved to hospital or anywhere else. I know a lot of people
do die in a hospital but if you’ve got the option I don’t think that’s a nice place
to die (crying). Not that there’s ever a nice place to die, you know what I mean,
but I would rather her be here. It’s just more clinical in a hospital… if she’s
comfortable it’s not like they can do anything. (Louise, daughter of Gwen)
154
While by no means perfect, the overall level of satisfaction with the regional RACFs
was high:
I don’t know how they could actually improve it here. It’s about the top of the
tree, the nurses are all great, everyone associated with the [facility] are all
great. (Pat, husband of Cheryl)
6.6 Out-of-region inpatient hospice as the place of death
For this discussion, ‘hospice’ refers to the specialist multidisciplinary palliative care
inpatient unit across the border in Canberra, Australian Capital Territory, 115
kilometres from Cooma, and 200 kms from Bombala. None of the interviewed patients
identified the hospice as their preferred place of care or death, although Marianne and
George considered it an option. In order to gain a perspective from at least one FCG
whose family member had died in the hospice, three FCGs were identified from
obituaries in the local paper. One bereaved FCG, Andrew, responded to my request for
an interview.
For Andrew’s wife, the out-of-town hospice was her preference. From Andrew’s
perspective it was because Elaine felt safe there. She had complex pain issues and over
the years this had been poorly managed by both the local hospital and the tertiary
hospital in Canberra. Prior to her death, Elaine had been admitted to the hospice on two
occasions for pain management. This was a place she trusted and believed it was the
only place that would keep her comfortable:
She didn’t have to worry about anything. It was all pretty much taken care of.
She was back on pain management…she was still in a lot of pain, however, it
was a lot better than what things were. She was in a quiet environment where
she could, I suppose basically withdraw from the world, and there was always
one of us with her. And in [the hospice], if she was in need of anything you hit
the button and within a couple of minutes you could get it. (Andrew, husband of
Elaine)
155
Other patients and FCGs acknowledged the expertise of the hospice, in relation to pain
management:
I imagine that [the hospice] would do a much better job [in pain management]
because that’s what they’re devoted to, they’re focused on. They have palliative
care doctors and nurses. (Susan, daughter of Kevin)
[If I had major pain issues] I’d go to [the hospice]. I think I would just make a
choice to go where the specialists would keep me comfortable. (Marianne)
For most participants, the hospice was not an option, mainly because of its location out
of region. For one FCG, the concept of ‘hospice’ was a city institution where families
could hand over their responsibility of caring for their family member, “In Sydney,
you’d just disappear into a hospice.”(Amanda, daughter of Ivan)
Except for Betty (who had a distressing experience with her brother’s hospital
admission), and Elaine (who preferred to die in the hospice), the need for a rural hospice
was not identified, as most participating residents had confidence in the existing
regional support. Marianne, whose story will be told in the preface to chapter 8, was
ambivalent towards the hospice.
6.7 Congruence between actual and preferred place of death
All the participants identified a place they or their family member would prefer to die.
There was congruence between actual and initial preferred place for six research
patients. While I was unable to interview all the bereaved FCGs, there were legitimate
reasons for patients having to move from their preferred PoD. The views expressed in
the interviews suggested that the patients and their FCGs accepted the move, at the
time.
156
6.8 Concordance between preferences of patient, family caregiver, and
healthcare professionals
Consistent with the findings reported by Davies et al. (2016), there was good
concordance between patients and their FCGs around preferred PoD. For the patients
and their FCGs, the preferred PoD usually meant the place where the patient felt safe
and could maintain their identity, value and control. Initially this was their ‘home’ or
their private residence. Except for George, the patients’ and FCGs’ preferences were the
same, with the FCG making every effort to achieve the preferred PoD:
He would prefer to stay at home, he’s a home person … he loves being home, so
we try out best and use all the facilities and support to keep him at home as long
as possible. (Deedee, wife of Kevin)
In discussing the options for PoD, some FCGs indicated that if the circumstances
changed, for example pain became unmanageable, they would explore all the
alternatives. However, most FCGs felt the final decision to move from home would be
left to the patient: “I think we would just leave it up to [Dad]” (Susan, daughter of
Kevin). The GPs and nurses were more pragmatic. While wanting to respect patients’
choices, they saw their main role as controlling symptoms. The two GPs, palliative care
nurse, and community nurses all gave examples of incongruence between patients’ and
FCGs’ preferred PoC and PoD. In some situations the FCG accepted the patient’s
wishes, at detriment to themselves and the patient:
[Place of death] is important, but it’s not the only factor that I would think
about. And I’m not really talking about Ivan in this particular case, but quite
often people make these decisions and the families just can’t cope with it, and so
when the patient gets to a certain point the families, despite the fact that they
wish to die at home, the family are just completely falling apart, and you end up
bringing them into hospital. Often at that stage, the patient is so unwell and
clouded they’ve sort of almost lost contact with where they are, so then it
becomes a matter of almost treating the family rather than the patient, ‘cause
the patient’s not really aware anymore and sometimes the family are just not
physically up to the physical demands of having someone die at home, and
157
sometimes the family aren’t really behind it all, but they’ve just sort of
acquiesced to the patient. (GP)
6.9 Chapter summary
This chapter has described where the 15 research patients died. It has explored the
preferred PoC and PoD of the patients, their FCGs and HCPs. All the participants had
an initial preference. All the patients, including the RACF permanent residents, stayed
at home (or in their usual place of residence) for as long as was possible. Patients were
willing to consider alternatives to home for PoC, providing it was short term, to access
specific treatments (such as acute pain management or transfusions), and they could
return home as quickly as possible. Home (or UPoR) was the preferred place for routine
care regardless of the preferred PoD. Home was also the initial preferred PoD for most,
and while some were able to die in this place, the majority of patients and their FCGs
had to adjust to changing circumstances. Most accepted the change. The meaning of
home was explored through the experiences and perspectives of the rural participants.
Home is more than the material structure. The most significant features of home include
a place of security, control, memories, self value, belonging, relationship, and a refuge
from the outside world. If these elements could be found in places other than home then
the alternative PoD became ‘acceptable’, and the new preferred PoD.
The findings reported in this chapter contest the notion that a ‘home’ death should be a
measure of the success of palliative care. Some research patients died well at home,
while others died well in the local hospital, RACF, or hospice. Some patients died well
in their preferred place. Others, like George, did not die well in their place of choice.
Place of death is only one element contributing to a ‘good’ death. The following chapter
explores further the rural concept of the good death.
158
159
CHAPTER 7: THE RURAL ‘GOOD DEATH’
Ivan’s story
Ivan’s death, according to his GP, “was one of the most positive death experiences”
he’d been involved in. Ivan had been living with chronic leukaemia for 8 years or more.
Two months before I interviewed Ivan, his leukaemia transformed into an acute form,
with a very poor prognosis. Ivan was a 5th generation Monaro sheep farmer. He lived on
the family property, where he was born over 80 years ago. The property held many
happy, but also tragic, memories. Of his three adult children, two had moved away. His
son remained on the property, where he lived with his own family.
The farm was about 15 kilometres from town: close enough for Ivan to be actively
involved in his church and community, but far enough away from the hustle and bustle
of town. The farm house was cosy, and adorned with items that reflected Ivan’s love for
his family and his love of life. Ivan was embedded in the community and embedded in
the deep tapestry of his farm. Five years earlier, Ivan drew up his advance care
directive. His plans were to live a long, productive, healthy and happy life, and when
the time came, to avoid machines and futile life prolonging interventions, and to die on
the farm. His intentions had not changed.
Once his leukaemia transformed, Ivan was admitted to the oncology isolation ward at
the tertiary hospital. This was an unfamiliar environment. He longed to be home,
surrounded by nature and “all the trees.” Ivan was far from home, the place where he
could “paint and find peace with God.” Ivan made the decision that for him there would
be no more futile treatments in cold sterile isolation wards. He was going home.
It was a wet Monday morning when I interviewed Ivan. He was very frail but willing to
tell his story. There was a sense of calmness throughout the home. Ivan lay in bed
looking out the large double French doors that opened directly onto a secluded, shady
cottage garden. This was obviously a special place on the farm. A place to sit and relax
after a hot, tiring day on the tractor or chasing sheep, a place protected from the howling
160
snowy winds. The farm dogs were asleep on the mat outside the glass doors. We talked
about his illness, his hopes, his faith and his artwork.
It was obvious that Ivan could no longer care for himself. It took all his energy just to sit
up in bed to take a sip of water. He could still say his “prayers in the morning” but
could no longer “drag himself to church on Sundays.” His adult children had all come
home to care for their father. They had not lived together for many years, and with three
individual adult personalities, they found themselves tripping over each other and
becoming exhausted (physically and emotionally) in their attempt to do the best for their
father. They harnessed their collective strengths and “eventually worked out a roster”.
This allowed Ivan’s son, Paul, to continue working the farm, and the others to take turns
having ‘timeout’, to rest and gather their strength to face whatever lay ahead. With
limited energy and still “lots to do”, the family ensured Ivan’s energy was spent on
important things, like spending time with family and in the garden.
Team work was essential if Ivan was to remain home. A combination of friends and
health care professionals supported the family. Paul described “an endless stream [of]
Community Health troops” which at times became overwhelming. While each person
was keen to help, at times Paul felt there was a lack of co-ordination or a blindness to
“the overall picture.” Ivan was dying, he didn’t need dietary advice to build up his
strength. There were endless phone calls, and a steady stream of visitors who never
came empty handed. The kitchen was overflowing with homemade cakes, biscuits,
soups and casseroles. The community nurse visited. Fortunately for Ivan, his GP was a
family friend, so distance was no barrier for home visits.
Ten days after my interview, Ivan had a particularly restless night. He suddenly became
confused and agitated, most likely the result of a cerebral haemorrhage. The family
were exhausted. The girls needed to get away, to renew their strength, so took a drive in
the countryside. The GP dropped by on his way out of town. It was obvious to him that
Ivan was dying. Medications were arranged, but before the community nurse arrived at
the home, Ivan died. Ivan died at home, in his own bed, in his son’s arms. This is what
Ivan and the family had hoped for. This was a good rural home death.
161
Introduction
Ivan’s story illustrates that a ‘good’ rural death is possible. It is a powerful story of
family, home and community. Ivan’s death illustrates the meaning and importance of
home, as described in the previous chapter. For Ivan, home was a place of security and
control; a reflection of himself, his values and status; a place of permanency, continuity,
and a place to belong; a place of memories; relationships; the centre of activity; and a
refuge from the outside world. Having explored my participants’ preferences for PoD,
and the meanings they ascribe to these places, this chapter explores the position PoD
holds within the concept of the good death. Dying at home added to the ‘goodness’ of
Ivan’s death, but it wasn’t the only feature. Security, maintaining identity, self-value,
personal control and normality, and continuing to live while dying were contributing
factors. Absolute PoD is meaningless unless it is placed into context.
This chapter, explores what it means to die well in the rural Snowy Monaro region,
from the experiences and perspectives of a sample of the residents. I begin by drawing
out what is meant by the ‘good death’ among this population. The factors patients,
family caregivers (FCGs) and health care professionals (HCPs) consider to be important
in defining the ‘good death’, namely symptom control, place of death, the presence of
family, informal and formal support from the community, autonomy and spirituality, are
discussed. I describe the way previous experience moulds the expectations as to what a
‘good death’ should look like, and changes in perspectives over time. This chapter
concludes by exposing the core element of the rural ‘good death’ as described by the
participants.
7.1 What is a ‘good death’?
Ivan’s story illustrates a ‘good’ rural death from the perspectives of his son, who was
present at the time of death, and his GP, who was absent. What made it ‘good’? After
all, a beloved father died in the midst of acute agitation. There was no medical presence
at the time of death. The GP had left; the community nurse was away preparing
medications. For some participants in this study, the ‘good death’ related to the time of
death, for others it was the weeks leading up to the death.
162
A ‘good death’ might be defined by its rapidity:
The only good dying I can think of is quick (laughs). But possibly just being very
comfortable, no pain and just at peace. (Carol, wife of Ryan).
Despite his illness and weakness, Ivan was able to continue living while he was dying.
He was able to participate in his normal activities, albeit much slower and for shorter
periods of time. Except for the last few hours, he remained lucid. In the following quote,
his GP highlights the importance of maintaining physical and cognitive capacity until
the time of death:
It was fairly rapid and he had capacity, right up ‘til the day he died. He could
still get out and about and do things and I saw him at a family birthday, it would
have been probably…ten days before, and he was out and about and quite active
and visiting people and it was really only the last day that that stopped. So he
went from being sort of quite active, so in that respect, he didn’t have a
prolonged period of being bed-bound. But that was just lucky, I suppose.
(Ivan’s GP)
Ivan’s rapid death contrasted to Daniel’s experience with Leanne. Daniel struggled with
the length of time his partner was actively dying. For five days, Daniel sat at her
hospital bedside, as Leanne lay in a coma, wasting away, unable to engage in
conversation. Daniel’s concern was not only for himself but also Leanne:
The last four or five days like it… I guess it seemed as peaceful as it could have
been, I don‘t know. I guess if four or five days had of been two days it would
have been better. She didn’t get much joy out of just lying there. (Daniel, partner
of Leanne)
Other participants felt a ‘good’ death was a ‘timely’ death. Ivan was in his 80s.
Although he was still active, he had lived a long and productive life. He had children
and grandchildren to continue the family line. Other participants felt it important there
was time to prepare for death, to face death with a sense of life completion, as in the
following quote by a patient:
163
Knowing I’m ready, everything sorted out…I know we’ve all got to go so it
doesn’t matter, I’ve nearly made 75. (Ray)
Family caregivers reiterated this point about feeling there was no more “work to be
done”, with additional concerns that their relative not be overwhelmed by negative
emotions:
I suppose a good death for [Stuart] will be [if he can say] “I’m only 55, I have
only retired, yes the house and everything I worked on so hard, I won’t be able
to enjoy it; however, I enjoyed doing everything I’ve done. I don’t have any
regret. Of course I’d like to live longer, but I can die now, and fall asleep
feeling, gee there was so many good things I enjoyed”. So that would make a
good death. And that there’s no longer the anger. That there’s no fear. (Fred,
brother of Stuart)
I believe a good death, when the times comes, he knows himself and he can say
what he wants to say with his family. (Deedee, wife of Kevin)
The rural palliative care nurse made the comparison between a ‘good’ death and a
‘good’ birth, and commented on the differences between calm and tranquillity, with
everything going along as expected, to chaos, with nothing going as planned or as
hoped for:
I personally like to think that they are comfortable, peaceful, everyone around
is accepting, and that it’s a really nice transition from life to death, a smooth
transition…I see it along the lines of birth. If it’s a really traumatic birth and it’s
suddenly this rushing to theatre and a caesarean because things are going
wrong, that stays with the parents for, you know for a couple of months, maybe
they kind of retell that story, “Oh, it was so scary.” And I always think that
about death, too. If it can be a really lovely peaceful transition then that’s what
you remember. (Palliative care nurse)
For some participants, expectations matched reality, as demonstrated in Ivan’s story.
Before he died, Ivan’s daughter shared her hopes for her scenario of her father’s death:
164
A good death for me would be just to fall asleep [in the garden], in the chair in
the sun and … in a happy situation, like sitting in the sun with the sheep dogs
around him and just falling asleep in the chair. (Amanda, Ivan’s daughter)
Reflecting on Ivan’s death, Ivan’s son said:
I mean a death in his own bed and looking out his own window with his own
garden, with feeding the chooks two days beforehand, with a brown snake in the
garden, ‘cause that’s what brown snakes do in Australia and with all the family
coming and going. (Paul, son of Ivan)
The GP also concurred, that in his opinion, this was a ‘good death’:
Well I’d say, his death was probably one of the most positive death experiences
I’ve been involved in …. he died in his son’s arms on his bed, I mean, what
better way could you go? Like his son had his arms around him when he died, it
was quite an experience. I wasn’t there at that time, I’d been there just a few
hours before and I went there afterwards as well. (GP)
While the experience of death and dying was different for each participant, there were
common themes surrounding the ‘good death’. These themes were consistent with the
literature review in Chapter Two, and revolve around the western concept of a
‘controlled’ death. The physical, emotional, social, spiritual and cultural aspects of
dying were all identified by the participants, and are discussed throughout the following
section.
7.2 Factors associated with a ‘good death’
The participating patients, FCGs and HCPs reported a set of factors associated with
‘dying well’ within the Snowy Monaro region. Five major themes were identified that
represented concepts the participants considered important for a rural ‘good death’:
symptom control, place of care and death, family presence, community support
(informal and formal), spirituality, and maintaining identity and control (autonomy).
165
These themes all underpin the deep personal need to be valued, to feel secure, and to die
in a safe environment.
7.2.1 Symptom control
Good symptom control was identified by all the participants as essential to achieve a
‘good death’, and was probably the main facilitator. Most participants specifically
mentioned the importance of being pain free:
I think the last period as you’re leading up to death, I think, should be free of
pain. (Rodney, son of Barbara)
To be as comfortable as possible. It’s awful seeing [Dad] in pain and feeling
like you want to jump to do something and help but you know you can’t. (Susan,
daughter of Kevin)
Other participants spoke of the need to control other symptoms, as all symptoms, no
matter how distressing, have an impact on patients’ quality of life. Ivan’s recent swollen
ankles meant more than just physical discomfort. To maintain dignity, he always
dressed well when going to town. While he could wear slippers around the house, this
was not an option in town. His daughter reflected on the consequences of this new
symptom:
I don’t want him to suffer. Like today he’s got puffy feet, that’s new. And the
implications of that is the shoes don’t fit. And if the shoes don’t fit, we can’t…
well we can’t go to town because that’s quite undignified. (Amanda, daughter of
Ivan)
Pain and other distressing symptoms were feared, not only by the patients, but also the
FCGs. The following FCG spoke openly of how her partner’s breathlessness distressed
her and how she felt uncertain about best to support him:
Clare: Seeing him not being able to breath, you can see it frightens him. It would
frighten anyone. I mean hold your breath for five minutes and you can see that.
Interviewer: And it frightens you too?
166
Clare: And me of course, yeah. We sit there and he tries to catch his breath, and
I sit with him. You know, he sits on the reclining chair, that’s where he sleeps as
well. And I say “take some deep breaths” and… I sit with him and rub his
shoulders and he goes, “I’m right now”. But I don’t know if he is. (Clare, wife
of Ray)
One of the features that helped Ivan to die well was that he maintained a good level of
cognition right up until the last few hours. Patients feared the loss of awareness and
decision making capacity. However, a relative might be willing to accept some loss of
cognition if it meant symptom control could be achieved, as in the case of Stuart:
So I’m hoping to see [Stuart] in an acceptance, peacefulness, no pain of course.
Well I don’t think there will be, from what the staff have told me of his condition.
And at the same time, it’s a pity that if they have to give him so many drugs that
he just sort of passes out and dies in his sleep. But, I guess if he’s not afraid of
death he may see it as the next adventure, and that would be great. (Fred,
brother of Stuart)
The word ‘euthanasia’ was not used by any participants, nor was the request for
physician assisted suicide made by any patient or FCG. However, for some FCGs, such
as Nathan, good symptom management implied not prolonging suffering:
I asked them to give him some more pain relief, but that was just as much to
hurry up the process as anything else, I couldn’t see any point him lying there
possibly in pain, you know, sort of making groaning noises and that. (Nathan,
son of Gordon)
A ‘good death’ meant that life was not prolonged unnecessarily by futile interventions.
Peter, Marianne’s husband, had watched his wife go through many painful treatments,
and while he was grateful for the extra time they had together, he questioned the
fairness in prolonging life:
Because the treatment is available it prolongs life, but it prolongs suffering as
well. (Peter, husband of Marianne)
167
Some FCGs spoke of their concerns for their family member based on past experiences.
One of Paul’s elderly relatives had suffered a protracted death following a stroke. She
had lost dignity; she had no quality of life. Paul did not want this scenario for his father.
Ivan had made it known on his final discharge from the isolation ward in Oncology, he
did not want any further futile treatments. Paul respected his father’s decision and
autonomy over his decision making:
I watched my elderly relative die…. She had a stroke and was unconscious and
she was rattling away full of morphine. And not only morphine but she was
dehydrated. I couldn’t believe, she looked frightful and I said to the young boy, I
now realise he was probably a young registrar, “what do you think you’re
doing? I wouldn’t do that to a sheep”…. And that was indicative of our
society’s inability to deal with death in a humane way. And that was the case of
because medical intervention could and medical intervention had kept her alive
for weeks as a vegetable. Yeah, and I thought that was cruel. (Paul, son of Ivan)
While most FCGs would prefer that their relative was symptom free, there was an
acceptance of compromise. Speaking of Marianne’s experience with myeloma, her
husband commented on her ability to adjust:
Over time she has come to, well learned to appreciate even though she’s unwell,
there’s still things she can do, relationships to be had and things to be seen
through. She has the personality that allows her to make the most of the
situation, whereas another person might have got depressed or given up or just
lashed out or whatever, through frustration, but she has been able to, she has
such an accepting way of coping that she’s been able to make the most of it,
which is a strange thing to understand because through daily discomfort and
pain and everything she’s still able to find joy and meaning in little things. She
has an acceptance of her fate which should, you know, help us as family to cope
with the whole situation. If the person is constantly fighting or afraid to die it
would be terrible because just that anxiety has to be taken on. (Peter, husband of
Marianne)
168
Patients also demonstrated their ability to compromise and adjust to the symptoms as
the disease progressed. In the following quote, Gwen adopts a pragmatic approach
towards her planned whole brain radiotherapy, seeing in her experiences something that
could benefit others in future:
You know you’re going to get worse but you don’t think about that ….you get on
with life. I could sit here all day on my own and think “poor me” but I’ve never
thought “poor me” I just feel that there’s a reason for it and through what I’m
going through somebody else might be able to be helped. (Gwen)
Many participants equated a ‘good death’ with a peaceful death. This was more than
just being pain free. “Peaceful” implied the patient was comfortable, unaware of the
distressing situation, at ease physically, emotionally and spiritually:
Mum had a good death, 'cause she was in a coma, so she didn't feel
nothing…..just peaceful. That's all I can hope for my brother too, that he's not
in too much pain, and he just floats away. (Betty, sister of George)
Good symptom control was essential if someone was to die ‘well’ at home. For the
interviewed HCPs, symptom control was their main objective, and sometimes meant
compromising the respect for patient’s preferred PoD. After a difficult day trying to
control Marianne’s pain and restlessness at home (her story is narrated in the following
chapter), her GP indicated that if her symptoms had not settled she would have been
admitted to the local hospital in town:
If we hadn’t been able to [make her comfortable], she would have had to have
come into town. (Marianne’s GP)
Marianne’s community nurse, did not return to see Marianne settled, and was left with a
lingering sense of failure as a nurse, as she was unable to control her symptoms:
I saw her certainly at her worst…if I had of arrived 5 minutes before she died
and saw her peacefully dying, I would perhaps say “that was a really good
death” but my vision’s been skewed (laughs) by what I saw before…and that’s
169
put a lasting impression in my head. (Marianne’s community nurse)
7.2.2 Place of death
Place of care and death was identified as one factor contributing to a good death.
Preferences for PoD and the meanings ascribed to place were discussed in depth in the
previous chapter. Here I discuss how PoD and the congruence between actual and
preferred PoD fits into the concept of the good death. As reported in the previous
chapter, most patients and FCGs initially identified ‘home’ as the preferred place for
care and death; however, 10 of the 15 patients (excluding Barbara who considered the
RACF to be ‘home’) did not die at ‘home’. Along the illness trajectory patients
frequently had to move away from home. For some, the move reflected a change in
preferred PoD, for others the ‘new’ place was not of their choosing, but became an
‘acceptable’ PoD. In bereavement interviews with FCGs whose family member had to
leave home, it was apparent that the actual PoD did not have a negative impact on the
quality of dying. Marianne died at home, in her preferred PoD, however, her husband
regretted she did not die in hospital, where medical advice would have been
immediately available. He actually considered her home death to be a ‘bad’ death.
A home death was Marianne’s idea, it was not realistic….home was the right
place for Marianne…but it was too traumatic for us. (Peter, Marianne’s
husband)
In the initial interviews, most participants considered home to be a peaceful,
unproblematic place. However, some patients indicated they would consider moving
away from home if necessary. Frequently, patients who wished to die at home qualified
their preference for home with statements such as “if my symptoms are controlled” or
“providing the family are coping.” While patients had discussed their preferences with
their FCGs, it appeared not all FCGs were as willing to change their preferred PoD. As
an aside to this discussion, it is fitting I add the following personal reflection on the
situation that arose for Marianne and her family. One of the personal tensions I had in
conducting the interviews was that in guaranteeing confidentiality I was unable to pass
information between patients, their FCGs and HCPs. I had been involved in Marianne’s
care but was away when she died. Marianne had indicated in her interview that while
she wished to stay at home to die, she was uncertain if her husband and adult children
170
would cope with the physical and emotional responsibility of providing 24/7 care.
While she had told her family she wanted to die at home, she was actually very open to
the idea of receiving end-of-life care in either the local hospital or the Canberra hospice:
Say I was semi-conscious and they couldn’t control my bowels…anything like
that. If my family can’t manage that and I’ve told all the girls, and the boys too,
if it gets to the point where I need full nursing care, like the bed changed
because I’ve soiled myself or stuff like that I don’t want to really, I don’t think I
want to be home if we’re at that stage, just for the sake of dignity and I’d feel
uncomfortable… and you can’t make those decisions ‘til they’re upon you.”
(Marianne)
Peter’s determination to keep Marianne at home had a negative impact on his
bereavement. He tried desperately to find some positive feature, something he could
attribute to a ‘good death’ or a sense that he and the family had helped his wife to die
well:
The ambulance man said “it was fantastic to achieve [a home death]. Usually it
becomes too physically and emotionally draining, becomes too hard and most
people phone us to take the person to hospital.” I suppose we achieved
something significant, but it was extremely traumatic. I had to cope, I was the
leader. I had to be there for the family…We had a plan, I had to do everything to
keep to our plan. Marianne wanted to stay home and then she couldn’t tell us
what she wanted. (Peter, husband of Marianne)
Actual PoD is often a compromise between honouring patients’ wishes and being
mindful of FCGs’ capacity and ability to provide care and take on the responsibility.
Obviously, it is impossible to determine how important the congruence between actual
and preferred PoD was to deceased patients; however, as highlighted in the case of
Marianne, the FCGs must be considered as it is the FCGs who are left with the long
term memories of the death:
Marianne did want to die at home. But… her husband wasn’t quite as
committed to that, or he was committed to it, because of her sake, but that
171
wasn’t what he would have chosen necessarily, because he found that whole
caring aspect quite challenging I think… I mean he doesn’t have any sort of
health background and so physical care and that type of thing, and also the
emotional intensity of being with the one that you love without the backup and
support of medical people in your own home. (Marianne’s GP)
As alluded to in the previous chapter, the interviewed HCPs appeared much more
pragmatic when it came to PoD. While endeavouring to honour patients’ wishes, PoD
was not considered to be the major contributor to a ‘good death’:
With any patients, [home] is important, but it’s not the only factor that I would
think about. (GP)
Ivan, whom we met in the preface to this chapter, did die at home, and on all accounts
his was a ‘good death’. There is no doubt that home was very important to him and
contributed greatly to his ‘good death’. His identity, security and self-worth were all
embedded in his home and farm. Home was central to his relationships, activities and
memories. From information gained in the bereavement interview with his son, there is
no doubt that a hospital death would not have been ‘good’ for Ivan or his family:
I just keep hearing stories again and again that reiterate how lucky we were to
have a death like that. I haven’t heard of any deaths at home like that, most
people end up in hospital, wired up to machinery and far from their home
environment. (Paul, son of Ivan)
Most research patients did not die in their initial preferred place. The reader will recall
Gwen, whose story was told in the previous chapter. Gwen died in a RACF, the place
she least wanted to die. When Gwen’s daughter reflected on the quality of her death, the
PoD contributed to, but was not the overriding determining factor:
I think she died as well as she could have, if you know what I mean. To her, I
think, she would obviously have preferred to be at home, so it’s hard for me to
say. I think in the scheme of things, I think she did [die well]. (Louise, daughter
of Gwen)
172
Through the experiences and perspectives of the rural participants it appears that while
home is the preferred PoD for most people, preferences change throughout the illness
trajectory. Some patients died at home, their preferred PoD. Some of these deaths were
good; some were not good. Other patients died away from home, either by choice or
circumstance. Again some deaths were good; others were not good. While PoD is a
contributing factor to achieving a ‘good death’ it is by no means the most important
factor, and in some cases had no bearing on the quality of the dying.
7.2.3 Family
Many FCGs spoke of the importance of being present, if not at the time of death, then
the days or hours before. In the initial interviews, a number of FCGs spoke of their hope
of being present and their fear of being absent at the time of death. Being present would
provide the FCG opportunity to support and comfort their relative, and ensure they did
not die alone. Being present also provided comfort to the FCG in their bereavement,
knowing their relative was not abandoned and alone. In situations where the patient had
been agitated and restless seeing their loved one finally at rest, provided some comfort.
This was the last act of kindness they could offer their loved one:
I’d like to be there to say, “don’t be afraid, we’re here.” That’s what I’d like to
do. Be physically close to him… No, not on his own, and physically I can be
there close to him, and for him to hear, “don’t be afraid, don’t be afraid”…And
of course when he dies, his memory will, he’ll still be with us in our thoughts,
and every time we touch or see anything that we know he loved, we’ll be able to
recall him. (Fred, brother of Stuart)
He loved his family and his family was there, so it was a good death, yeah. We
had two little grandchildren, so dad’s great grandchildren [were also there].
(Nathan, son of Gordon)
One of the fears surrounding a hospital death was that the FCG would not be present at
the time of death:
My biggest fear is not being there when he goes. And if he’s in hospital there’s
a ten minute drive, lots of roos, traffic, so you’ve gotta try and get there in time
173
before he goes and my biggest fear is not being there, not being able to hold him
or hold his hand and telling him “it’s OK, I’ll be alright”. I think that’s my
biggest fear, not being there. (Carol, wife of Ryan)
In addition to being present at the time of death, some FCGs thought it important to
have opportunity and time to talk with the family. This would require good symptom
control and for the patient to retain capacity and awareness:
I think other things would be having family around, if that’s what you want. Like
the chance to say goodbye, for him to say goodbye, for us to say goodbye, to
express things you never said while you’re alive because you don’t need to say
it. Being able to say goodbye, to forgive, to reconcile if you need to, and to be as
comfortable as possible. (Susan, daughter of Kevin)
The importance of being present was illustrated by those who were present at the time
of death or had spent time with their relative before they died. A few FCGs lived out of
area and they all spoke of the importance of making the effort, and sometimes sacrifice,
“cause I more or less did put my life on hold” (Louise, daughter of Gwen), to spend
time with their relative before they died. A number of FCGs left their partners and
children, some still at school, behind at home while they came to the Snowy Monaro
region to care for their family member:
That’s one thing that I look back on and I don’t regret. That’s one thing I can
say, I spent the time with her…I don’t regret [leaving my children in Sydney]
when I look back on things, yeah, that’s one thing I’m pleased I did. (Louise,
daughter of Gwen)
It was very important that we were there, even though we weren’t there at the
precise time, but it was good that we were there and we could be there within
minutes after the event, but we’d been there all day. (Rodney, son of Dorothy.
After spending all afternoon with his Mum he had gone back to the motel for a
short rest).
174
I don’t want to go back [home to Sydney]. I want to stay [with my brother],
although I’ve just left everything in Sydney. But then I’ve got supportive
neighbours. That’s my community, it’s the neighbours and the church
community. But the neighbours keep an eye on the house and everything. So I
know I can stay. (Fred, brother of Stuart)
Not all FCGs were present at the time of death. For some this was devastating,
compounding grief and guilt. Betty did not want her brother, George, to die in her home
and had promised him she would be with him in hospital and hold his hand as he died.
Due to his agitated state, the nursing staff had requested Betty not visit, and so, George
died alone. In my interview with Betty, before George died, she articulated some hopes
that were not to come to pass:
George would want a quiet and peaceful [death], and to see everyone's face for
the last time, and then just drift off. I think that would be the best for George. I
mean not necessarily his friends, but his family… [The place] isn’t important
providing there’s love. (Betty, sister of George)
Elaine, wished to die in the hospice in Canberra, some 110 kms from home. Andrew,
her husband, was able to spend a lot of time with her, but on occasions had to return
home to attend to business. It was during one such trip home that Elaine died. Andrew
regrets his decision to return home that night, a wrong that can’t be put right:
Unfortunately, that night was one of the nights where I came back to sort things
out, and went back again. And I made it a short night, I think I left there about
ten o’clock at night, and I was back there at seven o’clock in the morning, but
she was already gone. So for me, I needed to be there that night. But I can’t
change that. I could have organised things better, I suppose. (Andrew, husband
of Elaine)
In the previous chapter, one of the interviewed GPs spoke about a recent hospital death.
The family was from out of town and were staying in the family room on the ward. The
man’s wife was present when he died; however, his adult children were getting dressed
in the family room next door:
175
I honestly thought that was a very good death… he didn’t want to come into
hospital, but they physically couldn’t have done it at home…I didn’t actually
think whether it was at home or at hospital made a lot of difference actually, it
was more about whether the family is there. Yeah, it was more about who was
there rather than where it was. Well his children were not with him the moment
that he died, but that didn’t seem to make a lot of difference, his wife was there.
(GP)
FCGs also play the role of advocate for the patient. This is a privileged position and
works well when “all the family’s in agreement and everybody knows what’s going on”
(Paul, son of Ivan). However, at times FCGs priorities differed from the patients’, as in
the case of Stuart and his sister:
Stuart told the palliative care doctor in Canberra, when I was sitting by the bed,
“I know my illness is terminal, and I want quality in whatever time I have left,
not quantity.” And that was difficult, because lot of the palliative care
interview… was [our sister] saying how she wanted a PICC line delivered. And
the doctor said, “well, there’s a high risk of infection with these [and he would
have to go to hospital] to control that”. And [our sister] had a few ideas…I
suppose she just feels she doesn’t trust people that they’ll look after him
properly…And I took Stuart’s arm, and I said, “What do you want to do, Stu?”
And he turned, and the doctor said, “Yes, a good question. Stu, what would you
want?” And that’s when he said, “I’d like to go home.” (Fred, brother of Stuart)
While the dying person is central to the concept of the good death, the level of care and
support for the FCGs must be acknowledged. If the FCGs are left exhausted, guilt
ridden or significantly distressed, complicated grief may result (Miyajima et al., 2014).
It is during the bereavement period that the FCGs reconstruct their impression of the
death. It is important that FCGs’ concerns, objections and reassurances are listened to
and validated. Some patients intuitively knew the thoughts and concerns of their FCGs,
while others needed to hear them spoken by their FCGs. Most patients’ decisions
regarding their care were influenced, to varying degrees, by the viewpoint of their
FCGs. A successful home death required teamwork:
176
A person can’t just have that right [to die at home] without the cooperation of
their loved ones. (Community nurse)
An additional issue for patients and their FCGs was raised by a community nurse. It is
difficult to know whether Marianne and Peters’ expectations were due to poor
communication or their desire to ‘hope for the best’. They had been told by Marianne’s
haematologist that in the terminal phase she would “slip away peacefully, and without
pain,” despite having suffered with complex pain throughout her illness. This mismatch
of expectations and reality had a negative effect of Marianne’s dying:
I think she possibly did not have a ‘good death’….because the family had
unrealistic expectations of what was going to be happening as she got further in
her illness. I think that the unrealistic expectations set up for failure. When we
got to her last few days, because it wasn’t what the family had been led to
believe by the treating specialist, there was some anger there, which I think
made it difficult for the family to fulfil a caring role, because there was so much
anger there…And I think that led towards a situation of panic, to be truthful. We
did manage to have her pain relatively well controlled, but there was a
restlessness and agitation, and when she was demonstrating that restlessness
and agitation she wasn’t in a fit state of mind to be able to express herself
verbally and that made the situation harder for the family, as well as the nurses
going in. (Marianne’s community nurse)
It inevitably falls upon the rural community to support the FCGs, although distant
family and friends may be available to contribute. The support is often equally shared
amongst informal and formal networks. Some of the research patients had limited
family within the region, so their support was mostly provided by the community.
7.2.4 Community support
In order for someone to have a ‘good’ death, the patient and their FCGs need to feel
supported. Gwen’s daughter felt that the large number of friends who visited the RACF,
helped Gwen to accept the placement. Gwen had been an active member of the
community, so visitors were familiar and ensured she did not feel isolated:
177
She had so many people that popped in and that was a really positive thing
about her being here too, is that people could do that…if she wasn’t mixing with
people I don’t think she would have been as accepting of the situation. (Louise,
daughter of Gwen)
Daniel also appreciated the support he received, not only from the local community, but
also Leanne’s international and interstate friends:
I was just amazed. But yeah [a friend] from the Sunshine Coast and [friends
from Canberra] and you know different ones that… spent time here over the
eight to ten weeks. Just reflecting on it, it was a good time for us as well. I mean
I guess we formed a closer friendship and you know Leanne was here for most of
it and she was up for just as much fun while she could still have fun when they
were here and… and I guess everyone was making the effort to make it as good
a time as possible for her and… and yeah as sad as it all was, it was still nice
that we all went through that together. (Daniel, partner of Leanne)
Friends also offered practical support, especially for those who had stored up social
capital before they became unwell:
Elaine had a lot of friends in Cooma, and she did a lot of community work. You
know she was really well known by parts of the community, and those people
were queuing up to do something. I’d have to send them away all the time, there
were far too many people. You’d hear a knock on the door, and open the door
and there’d be a meal left there. And that sort of thing happened many times.
People turned up one day and said, “Right, we’re going to clean your windows
for you,” and they went through the house and cleaned all the windows, you
know. It was amazing. (Andrew, husband of Elaine)
All of the participants spoke of the essential support required from health care
professionals in making a ‘good’ death. This support was not just medical advice and
management, but also came in the form of friendship, practical support, a listening ear,
and often just being there:
178
I think for someone to feel they’d had a good death would mean that they felt
supported through every step of the way, whether that means they’d chosen to
die at home or in a facility or in a hospital. (Community nurse)
7.2.5 Autonomy
Autonomy, and respect of patients’ choices, were cited as factors contributing to the
‘good death’. Most FCGs concurred with Susan (daughter of Kevin), who considered it
her responsibility to “respect [Kevin’s] wishes as much as we can.” Many patients
described how the scope of their autonomy narrowed as the disease progressed. Early in
the illness, most patients indicated it was important to participate in decisions regarding
treatment options and place of care. For Gwen it was important to undergo surgery in
Sydney.
All of the participating patients and FCGs expressed their personal preferences, wishes,
goals and priorities. One of the benefits of remaining home was to retain autonomy:
[At home], you can do what you feel you should do, whereas if you’re in an
institution you’ve got to abide by the rules of that institution. (Gwen)
For some, the fear of losing control caused much distress, “I’ve lost control [of my
body, the disease and my life] and I’m an emotional wreck” (Marianne). Likewise, the
disregarding by HCPs of patients’ and/or FCGs’ wishes, was also met with frustration
and disappointment.
While most patients and their FCGs accepted the move from home, as death
approached, participation in the decision making was important. Gloria accepted the
move to the aged care facility, providing she could continue with physiotherapy. Once
settled in the facility, Gloria and her daughter rejected the recommendation that she
move to a room closer to the nurses’ desk:
I asked Mum and she didn’t want to move. She’s already had to come to
hospital and then here [aged care facility]. She’s just got this room, and I think
moving her again is not really fair, especially as she doesn’t want to go. [Louise,
daughter of Gwen]
179
For others, respect for their decisions regarding the final PoC and PoD contributed to a
‘good death’. It was Barbara’s decision to return to Cooma, despite it being
inconvenient for her son to visit, “[Sydney] was getting her down, and she says, “I want
to get out of here” (Rodney). Ivan’s decision to leave the oncology isolation ward and
return to the farm, was respected, as was Gordon’s family decision to transfer him from
the tertiary hospital ICU to a ward bed in the rural hospital.
For some FCGs, the quality of dying was negatively impacted by the pursuit of
respecting patients’ choices. While some FCGs indicated they would leave the decision
making to the patient, “I think we would leave [the decision to go to hospital] up to
him” (Susan, daughter of Kevin), some were not given this opportunity as patients
became incapable to make decisions. Marianne’s family were intent in respecting her
previously stated wishes to remain home. As will be discussed in the following chapter,
according to Peter, (husband of Marianne), “her ideas were never realistic… I had to do
everything to keep to our plan.” This sentiment was supported by her community nurse,
and unfortunately, “the unrealistic expectations set up [the family] for failure.”
While patients, FCGs, and HCPs preferences were often at odds with each other, and
preferences changed over time in pace with changing circumstances, a sense of
autonomy was maintained providing patients were listened to, and participated in the
decision making.
7.2.6. Spirituality
Two patients, Gwen and Ivan, found peace in their Christian faith. While Ivan was no
longer able to attend church, daily prayers gave him a sense of contributing to others’
well-being:
If you just sat there and watched TV all the time, you're not doing anything for
anyone. So this is where prayer is important. The more you keep mixing, and
finding people that need prayer, and some of them need incredible prayer just to
keep going. And the more you do that, even if you're just a vegetable, and you're
in bed praying for other people…you’re contributing. (Ivan)
180
Gwen’s faith provided her with a source of spiritual sustenance and strength. Her
‘worldly’ suffering was a temporary process to be replaced after death with something
much greater and more peaceful:
If I didn’t have my strong faith I would be a screaming, screaming heap. I know
where I’m going and that’s the main thing…Peace, I have peace, I don’t need to
worry about suffering or what lies ahead…one day at a time. My faith gives me
a different view, a different outlook on suffering… because I know where I’m
going. (Gwen)
The concept of ‘worldly’ suffering (physical, emotional and spiritual) was also raised by
Kevin’s daughter, Susan. Her view of a ‘good death’ was associated with the trust in,
and assurance of an afterlife:
I think a good death is one where you have hope beyond death, where even in
the midst of your suffering and pain you know that actually death is not the end
but the beginning of actually the better life, the way life was meant to be. Life in
this world is full of pain and suffering, my dad is facing that at the moment, I
think the saddest thing for me is if suffering continues [after death] and there is
no kind of reward. (Susan, daughter of Kevin)
Most participants did not disclose any specific religious faith. However, spirituality did
not just relate to religious beliefs. Marianne found music to be a spiritual experience:
I don’t have any fear but I feel a form of grief. I listened to Brook’s violin
concerto yesterday, it was on the ABC proms concert and I just wept and sobbed
through the second movement…because it triggered how grief stricken I am and
how emotional I do feel about this part of the journey. At times I think I have
come to terms with it and then something like that happens and then I feel very
vulnerable and extremely sad (cries). We’d spent the morning with the
grandchildren and they’re so full of life, and you know I can enjoy all that
without feeling grief while I’m there. It’s music that triggers things for me, it’s
always music that triggers me. [It’s very spiritual and a] highly emotional,
emotive thing. (Marianne)
181
7.3 The central theme of the ‘good death’
The ‘good death’ for these 35 rural residents was influenced by adequate symptom
control (physical, emotional, psychological and spiritual), place of care and place of
death (not necessarily home), the presence and support of family and the community,
accessible health care professionals and other forms of formal support, spiritual strength
and comfort from faith beliefs and other mediums such as music and nature, and
maintaining some degree of control over the dying ‘journey’. After synthesising the
preceding themes, the central theme on which the ‘good death’ is founded can be
summarised in one word, SECURITY (Figure 7.1). A dying person is vulnerable, and in
most cases has lost control of their physical body. The disease has taken control. If they
can die in a safe physical, emotional, social, and spiritual space, and the FCG feels they
have helped create and protect a safe space for their loved one and themselves, then this
is a ‘good’ death.
To create a safe space, the patient and FCG need to take back whatever control they can.
As the literature review illustrated, a ‘controlled death’ implies control of symptoms,
timing and place. This necessitates autonomy and a right to decision making. Autonomy
respects identity and self-worth, and maintains dignity. Maintaining relationships with
family, friends, HCPs, and the community protects the safe space; but relationships can
also create disharmony. In a safe space the patient has the freedom to be who they are -
be it the old familiar person or the ‘new’ person created through the illness. For most
patients ‘home’ was seen as a safe place; but for many participants, the home quickly
became unsafe. All these research patients, except Marianne, supported by their FCGs
and HCPs, willingly moved from the ‘unsafe’ place to a safer place.
182
Figure 7.1: Interaction of themes and the creation of the safe ‘good death’
7.4. Chapter summary
This chapter has explored the concept of the good death in the Snowy Monaro region,
through the experiences and perspectives as narrated by 11 rural patients, 15 FCGs, two
GPs, one palliative care nurse, two community nurses and one RACF director of
nursing. While each stakeholder has an impression of what a ‘good death’ may look
like, the ‘good death’ is in the memory of the survivors, especially the FCGs. The ‘good
death’ is illusive, and has many different meanings. Many of the contributing elements
are not uniquely rural, but are shared across all settings.
For some participants the ‘good death’ referred to the moment of death, and for others,
the dying process. Maintaining a sense of normality, and continuing to live while dying
was important, and achievable, for many participants. Good symptom control was
voiced by all participants to be essential for a ‘good’ death. Symptom control included
management of physical, emotional, psychosocial, and spiritual distress of not only the
patient, but also the FCG. Place of death was one cog in the ‘good death’ wheel, but not
the ‘hub’. While home was the initial preferred PoD for all the participants, a death
away from home did not imply a ‘bad’ death. The presence and support of family, both
during the dying process and at the time of death, contributed to a ‘good death’.
However, when the FCG perceived or experienced a lack of support for themselves, the
death was sometimes constructed to have been a ‘bad’ death, even if it occurred at
Family
Place
(Home)
Community
Identity
Memory box
Security
Safety
Control
Autonomy
Symptoms
183
home. Community support, both formal and informal, helped the FCGs to feel
supported, and helped to maintain normality for the patients. A ‘good death’ was often
pictured as a ‘peaceful’ transition from life to death. However, for some participants, a
‘good death’ was not possible, with some settling for a ‘good enough death’
(McNamara, 2009).
The participants’ narratives illustrate that the key determinant of the ‘good death’ is
security, and not necessarily dependent on any particular place of death. For some
patients, like Ivan, home was a place of security. He died literally in the secure arms of
his son. For others, home was not a place of security. In rural Australia, the meaning of
home may carry a deeper significance and connection, particularly for people who have
lived long on the land. However, isolation and limited resources may make home less
safe than in urban settings. The following chapter explores the influence of rural
residency on the ‘good death’ and decisions surrounding place of death.
184
185
CHAPTER EIGHT: THE INFLUENCE OF RURAL
RESIDENCY ON THE ‘GOOD DEATH’ AND
PLACE OF DEATH
Marianne’s story
Marianne was married to Peter. They were not ‘locals’, having both grown up in the
city. They preferred the tranquillity of country living. They felt safe in their relative
isolation. Twenty years ago they moved from the city, with their seven young children,
to the farm. Twenty-five kilometres from town was close enough to be involved in
community life, but far enough away to protect their privacy. They were a self-
sufficient, independent family. Although not visible, there were neighbours nearby who
would respond to any call for help. There were patterns to life that created security.
Marianne was the lynch pin of the family; the organiser, the matriarch. Like many of the
children on the surrounding farms, Marianne’s children had been home schooled, so
home held many happy memories.
Marianne was interviewed in their huge rambling farmhouse. The last two kilometres to
the house was on a muddy, and at times, slippery dirt road. It was a lovely winter’s day.
The home was cosy; the wood fire in the corner burnt brightly. We sat in a new
extension that had been purpose built to ensure Marianne had a cosy place to sit during
her final months.
Marianne had been unwell on and off for over five years. She had the type of cancer that
cycled through response to treatment to relapse. This cycle continued over the years.
Bone pain, at times difficult to control by the local GPs, had been her constant
companion. The palliative and community nurses had visited Marianne at home
regularly over the five years. While required to travel to Canberra to see her specialist,
for the first three years she had received treatment locally in the five-chair, three-day-a-
week oncology outpatient unit. Her specialist minimised Marianne’s visits to Canberra
by being available by phone. Marianne had loved the local oncology unit; it was small
and personal. She appreciated the support she received from the oncology nurse and
fellow patients. They were all locals and Marianne felt a connection to the unit. Part
way through Marianne’s illness the oncology nurse became ill and resigned.
186
Unfortunately, Marianne had issues with the new oncology nurse and no longer felt
comfortable attending the unit. Marianne and Peter ‘wore a track’ on the road between
home and Canberra, a 270 kilometre return trip. The private oncology clinic in the city
was adequate, but did not have the personal country feel. In the beginning, regular trips
to Canberra were acceptable, even a “nice day out”. However, when her pain escalated
the journey became unbearable, but had to be endured as her symptoms and medication
regime were too complex for the local GPs and rural hospital.
A few months before I interviewed Marianne I had visited her at home in a professional
capacity. She was in crisis. She was very unwell and not expected to last the week. Her
adult children came from all parts of Australia. The house was busy with people coming
and going. Despite being gravely ill, Marianne continued to direct her care and the
household from her bed. After three days, a blood transfusion was arranged in Cooma.
This was a miracle. Within a couple of hours Marianne, who had barely enough energy
to get to the hospital, was up walking to the bathroom. This had been a “dry run” for her
dying, some eight months later.
Marianne wanted to die at home. Home was a safe place. Marianne and Peter thought
this “dry run” had shown them what to expect in a home death. Remaining in their safe
place may not be the best option, and might be too stressful and burdensome for Peter
and the family. There were “too many people with no medical training” staying at the
farmhouse. Marianne didn’t want her children toileting and washing her. Maybe the
local hospital or the out-of-region hospice would be a better option, but that could also
be stressful for Peter. With this immediate crisis over, the family returned to their own
homes, but in their absence, the “kids felt they should be doing more.” The oncologist
reassured Marianne and Peter that with her type of cancer, the last days of life “would”
be pain free and peaceful.
When Marianne entered the terminal phase of her illness the family all returned to the
farmhouse. Three days before Marianne died her GP went on holidays. Before leaving,
he handed her care over to one of his colleagues, the husband of one of Marianne’s best
friends. The following day Marianne became restless and agitated. The community
nurse visited daily. She remained available on-call by phone overnight; after hours
home visits were unavailable. Staying on the farm also meant GP home visits were
187
limited. With a busy day in town, the GP was unable to just “pop in between patients”.
He was happy to be interrupted with phone calls but there were “no medical staff to
provide [face to face] reassurance and interventions.” Peter and the family were on
their own for most of the time. Peter felt “unsupported and alone.” He did not want
phone nursing advice or care workers. He wanted the presence of “qualified medical
advice.” He was overwhelmed by Marianne’s increasing restlessness and agitation. This
was not what he had expected and it had been going on for two days. On the third night,
the night before Marianne died, medical support was finally available. After finishing at
the surgery, their GP friend went to the hospital, where he worked as a visiting medical
officer. He collected a syringe driver and medications and drove out to the farmhouse.
He stayed with the family until Marianne was settled. In the early hours of the morning
the GP returned to his home. Marianne died a few hours later at home, on the farm,
surrounded by overwhelmed, exhausted and distressed family.
Introduction
Marianne’s story contrasts to Ivan’s ‘good death’, yet they both died at home on their
respective farms. From Peter’s perspective, Marianne’s death was not ‘good’. Due to
relative isolation and limited resources home became an unsafe place to die. Marianne’s
story illustrates the influence of rural residency on the ‘good death’ and place of death
(PoD). The chapter begins with a description of the patients’ place of residence. This is
followed by a description of the participants’ perception of themselves as ‘ruralites’.
The aspects of rural living that either facilitated (life style, and familiarity with the
community, location and institutions) or hindered (distance, and lack of specialist and
allied health services, and local expertise), a ‘good’ rural death and determined the
patients’ actual PoD are discussed.
8.1 Participating patients’ place of residence
As previously discussed, in the literature review, there are many definitions of ‘rural’,
and many different types of rural communities. Even in this study, ‘rural’ was not
homogenous, and included towns, villages and farms. Each participating patients’
residence was unique, either in terms of location or the living circumstances. Some
patients lived alone, and others lived with a spouse/partner or family member.
188
Consistent with the quantitative data discussed in Chapter Five, the patients resided in a
cross-section of locations and degree of rurality. The majority of decedents identified in
the surveyed population palliative deaths, and patients represented in the interviews,
lived in a town with GPs and a hospital (66% (91/138) of surveyed deaths; 60% (9/15)
of interviewed patients). Farmers accounted for 17% (24/138) of surveyed palliative
deaths, and 33% (5/15) of patients identified in the interviews. Of the 15 patients
identified in the interviews, seven lived in Cooma (one lived alone; four with their
spouse; two in a RACF); two lived in Bombala (one lived alone in a private dwelling;
one in a RACF); one lived in his sisters’ ‘granny flat’ in Delegate; and five lived on
farming properties located between 10 and 38 kilometres outside Cooma. Three of these
farms were on bitumen roads; two were accessed by dirt roads for the last two and five
kilometres, making travel difficult in wet or snowy weather. All but one farmer lived
with their partner/spouse; Ivan lived on his own, but had an adult son living in a
separate house on the property. Bruce lived with his frail, elderly wife, but also had
adult family members living independently on the property.
8.2 Rural identity and resilience
The literature argues (Wilson et al., 2009a) that rural people define themselves as
different to urban residents, and this was supported by a comment made by Cheryl,
when comparing nurses in a Sydney hospital with the nurses in her RACF, “I’m from
the country … these are my sort of people”. Amanda and Paul, the adult children of
Ivan, made the observation that there is no ‘universal’ rural characteristic and that ‘farm
residents’ are different from ‘town residents’. We were sitting at the kitchen table,
drinking tea and munching on the neighbour’s homemade biscuits. Ivan was asleep in
the adjoining room. We were talking about receiving end-of-life care (EoLC) in the
country. Paul suggested that I needed “to define rural as in living on a farm or living in
a country town as being two separate things.” As children growing up on the farm, Paul
and Amanda were aware that they were ‘different’ to the town children:
I went to [the local high school], there was a big difference between the country
kids and the ‘townies’… they were just different. (Amanda, daughter of Ivan)
189
While Amanda found it difficult to explain the exact differences, Paul described how
‘farm’ residents are more independent, resilient and more adept at problem solving. This
resilience (flexibility and the capacity to recover quickly from adversity) and self –
reliance (the ability to problem solve rather than relying on others) was not necessarily a
personality trait but more out of necessity, born out of personal experiences and
expectations:
People on farms, by the nature of our business, we have to be able to fix
problems, whether it’s a sick animal or a broken pump or busted tractor or a
bushfire or fence or whatever it is, financial stuff, dealing with business people,
we’ve got to be able to do all those things. And we’ve got to be able to fix a
pump with our pocket knife, if that’s all we’ve got, that’s what it takes. Where a
town person would ring the mechanic and say “you fix it” and they’ll go to the
doctor and say “you fix that”… So it’s that sort of practical decision-making
that I think country people are probably more adept at than town people. (Paul,
son of Ivan)
Rural resilience, and in particular the resilience of farmers, was further described by
Paul when I interviewed him after Ivan had died. Paul described how farmers are faced
with new challenges every day. They are used to multi-tasking and finding novel ways
to solve multiple problems at the same time. The following quote was in response to my
question, “Is there a resilience?”
Absolutely, it’s about problem solving, so every time we walk out the door we’ve
got to solve problems. So this morning’s problem… we’ve got shearing next
week…it’s about to rain, the tractor’s been left up there [in the paddock], so
we’re going to have to walk up there today… we don’t have to walk, but we’re
going to walk five kilometres to go and get it, and we’ll check some sheep on the
way. That solves [two] problems that need addressing. (Paul, son of Ivan)
8.3 Challenges of rural residency
This section discusses the factors associated with rural residency, as narrated by the
participants, which impacted negatively on the ‘good death’ and PoD. During the
190
interviews, each participant was asked if patient care or FCG support had been
compromised due to rural residency. Bereavement interview participants were also
asked what influence, if any, rural residency had on the quality of dying and the actual
PoD of their family member. Almost universally, the immediate response was ‘no’, as
illustrated by Gwen and Ivan:
At the moment I can’t see any disadvantage living in a rural area…people are
kinder, people seem to rally round more, if they can help in any way they will
help. (Gwen)
No [disadvantages]. Quite the opposite. (Ivan)
However, when given time to think about the question most participants identified at
least one disadvantage, with the majority citing distance and travel. The following quote
by Marianne illustrates her change in perspective when given a couple of minutes to
consider the disadvantages of rural residency. Both Marianne and Deedee highlighted
the consequence of distance when illness prevented further travel - they would lose
contact with their specialists:
I honestly can’t think of any [disadvantages]…. Actually, a negative I hadn’t
actually thought of, the travel you have to put in…what I’m thinking as a
disadvantage, if the pain is too much to travel to Canberra…then I’m out of
touch with my specialist. (Marianne)
If he isn’t strong enough, it is only the travel. Otherwise I don’t feel any
disadvantages because I feel there is very good organisation and support here
[in town]. (Deedee, wife of Kevin)
8.3.1 Distance and travel
As in Marianne’s case, all the patients interviewed had received their initial diagnosis
locally; however, they had to travel out-of-area to Canberra, Bega or even further away
to Sydney, for further investigations, definitive diagnosis or treatment. Gwen and
Cheryl travelled to Sydney for some of their treatment. For Gwen, it was by choice, to
be near her daughter during and after surgery. Cheryl’s sub-specialist treatment was
191
unavailable in Canberra. While both were initially accepting of the need to travel over
500 kms from home, ongoing treatment and monitoring in Sydney was not sustainable
long-term:
I couldn’t travel to Sydney all the time because I run [my local business]…We
used to get letters to go and visit [the Sydney specialist]. People in the city think
you only live around the corner. Sydney hospitals might be good in some
respects, but for country people it’s not possible to get there. (Pat, Cheryl’s
partner)
While travelling was mostly acceptable in the early stages of disease, over the course of
the interviews it became apparent that the issue of distance and travelling became more
significant as the patient became sicker and more fatigued, when symptoms became
more troublesome, and when the patient entered the terminal phase. Travelling was not
just an inconvenience, it was also associated with physical distress. Travel for those
living on farms and in villages often meant a weekly trip to the nearest GP to obtain
regular medications. The following FCGs described the problems encountered:
And the distance, ‘specially when he’s sick, having to get up early hours and go
for appointments and scans and things like that and having to do the distance
was quite painful with the roads and the snow traffic… That can be quite
stressful. So that was a disadvantage …because it just puts everybody on edge.
(Carol, Ryan’s wife)
Oh, it was terrible. He’d have about two or three [pain tablets] on the trip in
and we’re only half an hour away, so it was terrible. (Betty, sister of George
who had to travel to the next town to see the GP)
Despite the inconvenience and hardship of travelling, many regarded travel for
treatment a price for living in a beautiful place. Early in Marianne’s illness her husband
considered the travelling a positive, as “a day out.” The interview participants included
those locally born and those who had moved to the region. Despite their origins,
patients and their FCGs equally accepted the compromise that needed to be made, as
illustrated by Bruce and Peter:
192
It was an inconvenience but if you live away from a metropolitan area, well it
appears that if you need certain things you have to go to a metropolitan area.
I’m willing to travel if [I have to]. (Bruce, generational ‘local’)
When you live in the country part of the accepted economy is that part of
income goes on travel, that’s just accepted…the quality of life we have here is so
excellent in many different ways, it’s a small price to pay. (Peter, husband of
Marianne, 20-year ‘local’)
On occasions, travelling for treatment was put into perspective by describing it in terms
of travel ‘time’ rather than ‘distance’. While rural distances to treatment may be longer
than in urban areas, due to faster speeds and less traffic on country roads, the time taken
is not necessarily greater. Marianne was discussing how ‘safe’ she felt living on the
farm, despite her isolation. She has family living in Canberra and compared travelling at
home to travelling in Canberra:
Here you’re 20ks but it translates to 15 minutes. In [the city], 5 kilometres
translates into 15 minutes.
Due to travel distances, contact with specialists ceased as patients became too unwell to
travel or to continue treatment. A number of participants (both patients and FCGs)
expressed a sense of grief over discontinuing their contact. One community nurse
associated the disconnection with a loss of support for the patient and their FCGs:
There’s distances often between where the treating doctors are and where the
patient is residing…and I think for someone to feel they’d had a good death
would mean that they felt supported [by their specialist] through every step of
the way. (Marianne’s community nurse)
It’s impossible now [to travel to Canberra to see the specialist], and there’s no
specialist here to come and see him. You know they won’t travel here. And that
upsets me to think, well he needs to see a specialist, but he [can’t] travel all the
way to Canberra. (Clare, wife of Ray)
193
It was an agonising decision… he wanted to have more treatment and we
weren’t sure if it was appropriate, and it was a [long trip] travelling
up…because his mobility was reduced. And I suppose if we had a visiting
oncologist here in Cooma, that logistic of travelling to Canberra to be reviewed
by an oncologist would mean just travelling [into town]. Well I’m guessing the
oncologist would not do a home visit. (Susan, daughter of Kevin)
The previous chapter reported that one factor contributing to a ‘good death’ was the
presence of family at the time of death. The fear of not being present was significant.
When FCGs had to travel to hospitals or the out-of-region hospice to be with their
family member, the travel distances became more problematic. While rural FCGs had a
strong desire to be present at the time of death they had to balance fulfilling their
responsibilities at home with maintaining a bedside vigil away from home. The reader
will recall, in the previous chapter, Andrew’s regret for returning home the night his
wife died in the out-of-region hospice. For Daniel and Leanne, the local hospital was
their preferred PoD and the fear of not being present was foremost in Daniel’s thoughts
every time he left the hospital to attend to chores on the farm, some 30-kilometres
away:
I guess I wanted to be [at the hospital] and then I’d come home and then I’d
think well it could happen in the next four or five hours …and so it’s hard to
come away and feel OK in case something happened. (Daniel, partner of
Leanne)
I did a lot of going backwards and forwards between [home] and Canberra, I
ran into problems when I stayed away from home too long, there were animals
and all sorts of things to be done, and things got completely out of hand. And
even to the point at one stage where a wild cat had moved into the house… I
would come home and do what I needed to do around [home] and fix the things
at home, and then gear myself up – the washing, etcetera, and that sort of thing
– and then head off. And then I’d spend another three days and two nights [at
the hospice]. (Andrew, husband of Elaine who died in the out-of-region
hospice)
194
Distance and travel impacted on most FCGs, regardless of whether they lived locally or
away. In the following quotes FCGs describe a number of scenarios they faced while
caring for their family member: the out-of-town daughter having to travel to Cooma,
leaving her own family; the family who had to travel from Cooma to be with their father
in ICU in The Canberra Hospital; the anguish of family trying to decide whether to seek
medical help locally or in Canberra; and the costs of travel when there was no family
home to stay in when visiting:
If [Mum] was in Sydney it would be more convenient, you know what I mean,
it’d be easier without the travel and stuff. (Louise, daughter of Gwen)
It was disruptive; we had to go up to Canberra every day for a month …It’s the
time, the fuel. Taking time off work, and it wasn’t just one person, it was my
brother and my wife and my kids and all that sort of thing, so we wanted [Dad]
back in Cooma…A lot of people say, you know, if you’re sick you should be in a
major centre … but I’m not really sure. (Nathan, bereaved son of Gordon)
The travelling hurt [Elaine] a lot. So, we would only head for Canberra
Hospital in the circumstance where we would consider that if we took her [to
Cooma Hospital] she’d probably be put in the ward and transferred to
Canberra Hospital anyway. The problem is that it would probably take a day to
[be transferred to Canberra], whereas if I just get in the car it’s only an
hour…but she has to put up with the pain in the meantime, not that she doesn’t
have the pain when she goes via the ambulance. (Andrew, husband of Elaine)
The tyranny of distance. Every time we went down [from Sydney to visit Mum]
we had to stay at least three days to make it worthwhile. We had to fuel the car
twice and to pay for accommodation plus food. We used to bring our own
breakfast material down, but we always had a hot meal out, and this was costing
us on average about eight to nine hundred dollars a trip. So [we] couldn’t do it
as often as [we] might have liked to. (Rodney, son of Barbara)
As discussed in Chapter Five, PoD was partly determined by where one lived and the
degree of rurality, with residents living on farms less likely to die in their usual place of
195
residence (UPoR) than those living in towns. This was in part a reflection of the
availability of, and access to health services. By Australian standards, none of the
participating patients were extremely isolated or remote. The furthest distance from the
nearest hospital and GP was 38 kilometres, although the distance to some part-time GPs
was greater after hours. The role played by distance and travel, in either facilitating or
hindering the quality of EoLC, was evident in the interviews. The health care
professionals (HCPs) indicated their availability to support patients dying at home, and
each other, became problematic the greater the travel distance and the more protracted
the terminal phase. The following quotes by a GP and a community nurse highlight their
perspectives on travelling distances to reach patients at home:
Kevin lived in Cooma itself, so he was five minutes away. So if there was ever
any issue or problem we were only five minutes away. Night time, [his wife]
seemed to be very capable…knowing that the hospital wasn’t far away, she
could phone the nurses at the hospital. I did go on-call the last two nights.
(Kevin’s community nurse)
If it had been a protracted illness it could have been quite difficult because he
lived 15 kilometres out of town and to get out there and back for nurses was
different to if someone was in town. (Ivan’s GP)
If it is a long way out of town, then it is really difficult. I remember we had this
young lady a few years ago up at [the ski resort], and that was really difficult to
access her. It was winter, she wanted to die at home…and I must admit we were
not encouraging her to die at home because it was wintertime…it was
snowing…and no doctor was going to home visit, and you kind of feel like
you’re on your own when you’re doing that. If you’re the one travelling to that
faraway distance, and [getting advice from the GP] all by phone, it’s difficult
for you to feel as supported as you do when you’re doing it closer to town.
(Kevin’s community nurse reflecting on a previous scenario)
For some out-of-towners, there was a mismatch between their view on the challenges of
distance, and those raised by the HCPs. This may reflect the out-of-towners’ acceptance
of daily travel. When I interviewed Marianne she downplayed her distance to town.
196
While she was still physically able to travel she was optimistic of the support available;
however, during her terminal phase her ‘15 minutes’ became an obstacle to receiving
support at home. The following sequence of quotes illustrates the different meanings of
‘15 minutes’. While some GPs were prepared to travel, the final quote illustrates the
restriction to travel due to responsibilities in town:
We’re only 15 minutes away, for acute care, for emergencies….we have the
help, we have doctors who are willing to come out, we have palliative care
nurses who are active and willing to come. (Marianne)
[Living on a farm] certainly made it harder, because you couldn’t just duck out
there. I mean I couldn’t just sort of pop out. If she’d been living in Cooma I
could have said between patients that I could go out for five minutes and just
have a quick look and see. But obviously it was a 40 minute round trip, so you
couldn’t do that. So her distance… made that aspect of it harder for sure.
(Marianne’s GP)
Because they lived out of town, it wasn’t like you could just pop in the car and
see how she was. You had to allocate time. If it’s raining that road can be
terrible …it was 20 minutes, 25 minutes to get out there… we don’t have an
evening service, so if we go out there it’s usually a long home visit because we
want to make sure we go through everything and make sure everything is in
place and everyone’s clear, so it would be a big chunk of your day. (Marianne’s
Community nurse)
The thing is I can’t always leave town because I’m on call for maternity…and
we’re supposed to be within ten minutes of the hospital… There are doctors that
have got anaesthetic call and the emergency doctor can’t get out of emergency
for their whole shift, so on the days where we’ve got ‘on call’ responsibilities
[we can’t leave town]. (Ivan’s GP)
The participants all eloquently narrated their perspectives on the “tyranny of distance”
faced by ruralites. However, distance was not the only factor negatively impacting the
‘good death’ and place of death.
197
8.3.2 Other challenges of rural residency
8.3.2.1. Lack of local specialist services
While travel distances prevented patients from attending their specialists, especially
once they became too unwell to travel, the lack of local specialists and interventions
were also considered a disadvantage of rural residency. However, as patients became
more unwell, they and their FCGs acknowledged that the need for further complex
interventions was no longer necessary, regardless of where they lived:
If a high level intervention would make a difference, you’re not going to get it
here. So Cooma can’t do platelet transfusions ‘cause of the short time between
extraction and delivery that’s required. But that’s the only technical issue we’ve
come across…No, the advantages far, far outweigh any disadvantages. (Paul,
son of Ivan)
Ready access to specialist services and expertise instilled confidence in some patients.
This was particularly true when patients had complex pain issues. A number of patients
spoke about the inability of local GPs to manage more complex issues. During a
hospital admission in Canberra, Marianne became quite fearful of being discharged and
returning home:
I got very insecure, which is very unusual for me cos I’m usually on top of
things, and I wanted to stay an extra couple of days [in hospital]. I was
frightened to come home because of all the drugs. [My specialist] said, “You’ve
got to go home, you’ll be right, you’ll be OK” and it was. (Marianne)
8.3.2.2. Lack of local medical expertise
While most had been satisfied with the involvement of their local GP at the time of
diagnosis, Susan talked about her disappointment in the delay in diagnosis of her
father’s cancer, which was not necessarily a rural issue:
The first [disappointment] would be 3 years ago when dad went to a local GP in
town, he had a lump, but it wasn’t really investigated [until] a year later. That’s
something we look back on as a disappointment, but of course that could have
198
happened with a GP in Canberra; I think it just depends on the particular GP.
(Susan, daughter of Kevin)
While most participants praised their local GPs, some felt they lacked the ability to
manage complex issues, such as complex pain and high dose opiates. The perceived
shortcomings were seen as multifactorial. Some of the inadequacies related to
education, knowledge and experience of the GPs. Some participants expressed their
concerns over the lack of funding and resources for rural hospitals, and the absence of
proper clinical handover by the specialists and tertiary hospitals. A number of FCGs
spoke of the lack of continuity of care between treating HCPs. Speaking from first-hand
experience, Andrew became quite angry as he spoke of the discontinuity in the care of
his wife, Elaine. Fred also voiced his frustration with the poor clinical handover:
The major issues that I can see are the lack of co-ordination and communication
between the services. And most especially in the country hospitals there appears
to be a big lack of funding… the support of the local GP is obviously critical. As
far as I can see, there’s very little support for a local GP. They need to be
across the sort of treatment that [their patient] had, and the history up to that
point, would help a lot. But it just seems to me that the local GP in many cases
doesn’t even get decent reports or anything. They’re not being told. And it’s
their patient. Their patient comes back and they have absolutely no idea what’s
happened in the meantime. (Andrew, husband of Elaine)
Canberra just sent him back to [the local] hospital with no proper medical
notes. They just sent him home to die, basically, without an adequate handover
… [The GP] said, “Well, we don’t have any notes from Canberra. We’ve got
nothing to tell us what’s going on.” (Fred, brother of Stewart)
8.3.2.3. Lack of allied health support
Limitations in accessing allied health support was cited as a disadvantage of rural
residency. Some FCGs lived in cities and commented on the paucity of, or absence of,
some services compared to more urban areas. In some circumstances, as voiced by
Amanda, this was just an inconvenience; for others, like Gwen, the limitations in private
home nursing reduced her chance of returning at home. (As an aside, since completing
199
my interviews, HammondCare Palliative Care Home Support Packages, funded by the
New South Wales Ministry of Health, are now available, providing up to 48-hours of in-
home end-of-life support by trained care workers (MacLeod et al., 2015)). One RACF
director of nursing (DON) felt the lack of counselling support affected both the patient
and the FCG:
And we’ve run out of glucosamine, so I went to the health food store, “oh we’ve
run out, we’ll order it.” I come in three days later, “oh we haven’t… the order
got misplaced, and we’ll reorder it.” And I thought welcome to country town.
Not that the glucosamine will make any difference but… (Amanda, daughter of
Ivan)
I think the family would have benefited from a counsellor or a social worker.
That would have made a huge difference, and then that would have had an
impact on Gwen, because she got upset when the family were upset. (DON re
Gwen)
While living only 10 kilometres from town, Carol spoke of the difficulty of leaving
Ryan on his own, and the difficulty of finding someone at short notice to sit with him.
While this did not prevent Carol from running errands, she was always mindful of the
time she was away from home:
If I had to leave him on his own I couldn’t go for very long…so that would be a
problem, leaving someone as sick as he is home on his own. You’ve gotta
organise [home care] in advance… you can’t just ring up and say look I’ve
gotta duck into town and do something. They can’t just come out, you’ve gotta
have that in place ahead of time. (Carol, wife of Ryan)
8.4 Benefits of rural residency
Despite the challenges faced due to rural residency, all participants stated the
advantages outweighed the disadvantages. One of the positives of rural living, voiced by
many participants, was the beauty and tranquillity of the rural lifestyle. Some of the
participants, like Ivan and Bruce, were born in the region; others had moved there.
200
Leanne had moved to the area only three years previously. Yet all participants loved the
region:
The advantages are really around lifestyle. (Andrew, husband of Elaine)
Leanne was just so grateful to be able to spend her last few years here … in the
peace [and quiet], with the animals, and just walk down the old road if she
wanted to. (Daniel, partner of Leanne)
Some participants, like Ryan, who spoke of the “good clean air”, believed rural living
had positive health benefits. Pat could not contemplate living anywhere else:
I would suggest you wouldn’t leave here (laughs), like you couldn’t get a better
place… as far as living and care goes you wouldn’t get it any better than here.
(Pat, partner of Cheryl)
A few months before Marianne died, she described feeling ‘safe’ in her isolation. It is
interesting to note that during her dying phase, the ‘isolation’ of the farm contributed to
home becoming an unsafe place for her family, and limited the availability of medical
and nursing support:
I just feel safe…the isolation I love, the fact that I feel really safe in my isolation.
(Marianne)
A number of common themes emerged from the interviews that participants
perceived to be a unique benefit of rural residency. Each theme facilitated a
‘good’ death and assisted the patient to die in their preferred place.
8.4.1 Personalised care
Many participants spoke of the value of being treated as an individual and receiving
personalised care. The concept of ‘being known’ and ‘knowing others’(Pesut et al.,
2011), highlighted as a unique feature of rural hospitals in Chapter 6, was a major
benefit of receiving and providing rural EoLC. Some participants commented on the
small turnover of rural staff that facilitated personalised care. ‘Being known’ gave Ivan
201
a sense of acceptance for who he was rather than being known for his illness. For Pat,
knowing the staff gave him confidence in Cheryl’s care. From Louise’s perspective,
Gwen’s acceptance of her transfer to the RACF was made easier because Gwen knew
the staff. It was not just patients and FCGs who benefited from ‘knowing’ as the work
of the community nurse was also enhanced by knowing the other service providers:
You know everyone. There’s a sense of freedom I suppose. Oh well freedom to
be the person you want to be. People accept you as you are. (Ivan)
[In the Sydney hospital] you’re only a number. Here you’re recognised as a
person. I can go home at night and sleep knowing [Cheryl’s] well looked after.
When she was in Sydney I’d be worried every 5 minutes wondering if she is
being looked after. So much turnover of staff in those hospitals up there, not
regular, they come from the agency and you never know who, never see the same
one twice. (Pat, partner of Cheryl)
And the fact that you know people, you’re working with people, so you know if
you need a service you know who you’re talking to. It’s not a stranger on the end
of the phone. The personal touch of a smaller town is really advantageous.
(Community Nurse)
However, ‘being known’ also had a negative side, as illustrated in Marianne’s story. A
conflict developed between Marianne and a staff member, forcing Marianne to stop
attending the local oncology unit. Before her illness, Gwen had been a volunteer in the
local RACF. One of the difficulties Gwen faced, after being transferred to the RACF,
was allowing other residents see her in her frail deteriorating state:
She didn’t ever leave her room, because she felt very self-conscious of the way
she was, because she used to see all those people and help out, you know, with
feeding and stuff like that…so I think that was in a way a bit detrimental in
terms of, she didn’t leave her room and…come out into the dining area. (Louise,
daughter of Gwen)
202
8.4.2 Informal community support
Participants spoke highly of the support received from members of their local
communities. Support came from neighbours, friends, shop keepers, other patients and
community members. Most of the participants believed the community support offered
was unique to rural communities and was a significant advantage of rural residency, and
would be hard to find in a city. It was this support that particularly helped the FCGs,
and was something they appreciated and remembered fondly:
I’ve lived in the area all my life, and contacts in Cooma I’ve had forever…. the
network spreads pretty fast. For example, one of the ladies in town took Leanne
to [a doctor’s appointment] one day, [she] just came up and took our car and
took her in.… They went and had lunch together and spent a nice afternoon
together when I couldn’t do it.… There were people there that were ready to put
their hand up for whatever… everyone just pitched in. I mean [one friend]
would have two lots of washing on the line every day. All little things, but
reflecting back on the very, very sad times but some good times had as well.
(Daniel, partner of Leanne)
So [a friend] dropped in with a fruit cake and [another retired nursing friend]
came up and gave dad five minutes of nursing care, which delivered more
nursing care than all the other nurses put together (laughs)…stick the pillow up
this way, get the feet this way and wash his feet ‘cause that will make you feel
better, or whatever it does. Basic nursing… and there were a lot of people came
to visit. (Paul, son of Ivan)
Informal support was mostly direct and practical, such as food, transport and home help;
however, some support was indirect, with many providers unaware of the significance
of a friendly smile or ‘hello’:
I think what was nice the other day, [Dad] was in [hospital] all day having a
blood transfusion, and people who know you walked past and oh, “hello Ivan”,
and they came and had a chat. It was nice to see friendly faces when you're in
hospital. Whereas you wouldn't get that in Canberra. (Amanda, Ivan’s daughter)
203
Just living in a town where you have your habits. You go to that shop to get your
coffee, you get to know people…and when you bump into them you have a
chat…. and in general you get to know the staff in the shop, it’s very comforting
in many, many ways. (Peter, husband of Marianne)
During Ivan’s illness, Paul was aware that dying was not just a family matter but
involved the whole community. In the following quote, Paul perceives the death of rural
residents to be more ‘dignified’ than urban dwellers. This is partly due to his
observation that rural people tend not to die in tertiary hospitals “hooked up to
machinery”. More significant is the community involvement, especially when the dying
person is a much loved and respected ‘local’.
Country people seem to have more dignified deaths because they’ve got a more
practical understanding of death. Now that could be for two reasons. One is
we’re rural types and our main [work is] to keep animals alive, and sheep only
go for five years and then we’ve got to get rid of them, so we’ve got that
practical animal aspect. But in the community sense, when someone dies or is
dying the whole community knows about it. When there’s a funeral, there’s a
heap of cars up [the road outside the cemetery] and the whole town stands
still… Sorry (cries)…And food, people bring us fruit cakes and soup and endless
food. And these bikkies are very nice… which is lovely. (Paul, son of Ivan)
The value of community was intensified after Ivan died. In the bereavement interview
Paul was eager to show me something. He was “holding a pile of papers three inches
deep” containing letters and cards from friends and acquaintances:
A lot of these got delivered on the day of the funeral. I know you’re asking …
“was he still part of the community in the last few weeks?”… But he really was
part of the community…a hundred and ninety odd people turned up to his
funeral. (Paul, son of Ivan)
For Marianne, informal community support was a two-way concept. While undergoing
chemotherapy locally, Marianne found mutual support in other patients. Inside the
oncology unit a special ‘community’ of oncology patients existed:
204
[It was] a communal thing…even meeting somebody new and going “would you
like to have a coffee next week?” Or sometimes you meet somebody you click
with and you actually want to continue a friendship, the cancer’s actually
brought you together and you don’t get that in the city. (Marianne)
When patients became too unwell to travel into town, the community nurse observed
that travel distances to farms did not deter friends from visiting:
They had friends that were coming to see them. I mean the distance [out to the
farm] didn’t seem to stop their friends, their friends were all very happy to come
out and see her. (Marianne’s community nurse)
Another informal support, not previously mentioned, were the local church leaders and
members. Three of the patients interviewed talked about the importance of their
Christian faith. It was important to Ivan that he continued to attend church; he liked
“singing all the hymns”. When that was no longer possible, the Pastor offered to visit
him at home to provide spiritual support:
Plan B for church was that the Minister could come out here. (Paul, Ivan’s son)
So there was a lot of church support people, friends of [Marianne] who came
and sat with her. (Marianne’s community nurse)
8.4.3 Formal support
Patients and their FCGs received formal support locally, and from specialist units in
Canberra and Sydney. There were a number of situations, possibly unique to the rural
setting, which assisted three patients to die at home. Unlike the majority of urban
medical practices, most rural GPs live in the community where they work. It is only
natural that many patients become friends. Strong ties develop between GPs and their
patients, and for some patients the GP will be more available than is the standard
practice:
I’ve never really practiced in the city, but I imagine that you don’t generally
practice in the area that you live, so where you’d probably be practicing, your
205
patient’s wouldn’t be people that you’d be socialising with, whereas, here you
do, and you sort of go out of your way, I mean, there’s patients’ that you’ve
known for many, many years who become sort of like friends anyway, and
when someone’s dying, I mean, that’s about it, that’s the end of your
relationship with them, so you do probably tend to invest more time just trying
to get them through that last little bit. (Ivan’s GP)
Of the patients interviewed, two were close friends with their GP, with a third patient a
near neighbour to their doctor. All three patients lived out of town and died at home
with the support of their doctors:
Well Ivan’s son is my best friend…it was a privileged position to be in…if he’d
had a protracted sort of pre-terminal phase at home, I may well have spent quite
a bit of time out there, that I might not necessarily have spent with another
patient. (Ivan’s GP)
Marianne was a close personal friend…I was very glad to be available for
her…and to be able to care for the family… I’ve never done that previously. I
don’t expect I’ll ever do it again. I think this was a oncer under these sort of
unique circumstances really...I mean I wouldn’t have done that for anybody else.
(Marianne’s GP)
Carol rang me at about eight [am] to say, “Can you come up and see Ryan, he’s
got a lot of pain.” And then when I arrived he certainly was extremely agitated
and restless, and I think had [the doctor, who was a neighbour] not been there it
would have been quite ghastly for me on my own, because we needed to quickly
do the drugs and get drugs on-board quickly, and [the doctor] was there writing
up the orders as we were doing it. (Ryan’s palliative care nurse)
One important formal support service that enable people to die well, regardless of where
they lived, were the palliative care and community nurses. If the need arose, the
palliative care nurse was available to home visit every resident in the Snowy Monaro
region. However, the frequency of her visits was dictated by where the patient lived.
Patients living in Cooma had the greatest access to the palliative care nurse. On
206
weekends, a community nurse was available to home visit. While home visits were not
available after hours, on occasions, when someone was close to death, the palliative care
nurse or breast care nurse were available for phone advice:
I never have feelings of fear or being too far away from help or anything… I feel
[help] is a phone call away, [the palliative care nurse is] a phone call away.
Even though I hardly ever [call] I know it’s there. I know a person with
myeloma in Sydney and he has no visits from palliative care nurses at all…So he
only goes to see the specialist, there is no support. I suppose if he asked, but
here we didn’t have to ask…here I’m overwhelmed by support, it’s almost too
much. The nurse would be here every week if I wanted her. It’s unbelievable,
and we live in the country! So to me the advantage is huge, you get immediate
care. Even the ambulance came within 20 minutes when we needed it once or
twice…like in the city it seems to take longer, so I think we have more than
adequate support and I love living here, I wouldn’t ever want to move.
(Marianne)
As in the case of the GPs, the nurses are also ‘locals’ and part of the community. The
advantages of ‘being known’, and continuity of care enhanced a sense of trust in the
nurses. For some of the patients interviewed, the same nurse had been visiting and
providing support for a number years, often from the time of diagnosis. Andrew’s wife,
Elaine, had received all her chemotherapy locally. He commented on the flexibility of
the local nurses, compared to the urban nurses:
[The local oncology nurse] her knowledge and flexibility when things weren’t
working right, her sheer knowledge of the thing, and ability to work around
things and get things going, is way above what we found in Canberra, where
they have the new ‘you beaut’ facilities. (Andrew, husband of Elaine)
8.4.4 Convenience of rural residency
Despite the travel distances, once in town, there are conveniences often not found in
cities. When FCGs are ‘time poor’, the convenience of parking at the door, and lack of
traffic jams was reported as a benefit of rural residency:
207
In a small town you know exactly where everything is, if you need something you
know exactly where to go, shopping is not difficult, you’re never held up, you
can always park, there’s so many practical pluses in living in a place like
Cooma. (Peter, husband of Marianne)
8.5 Rural residency and safety
In the previous chapter, the central theme of ‘safety’ emerged as the main determinant
of the ‘good death’. A ‘safe death’ was one that occurred in a safe place. In illustrating
the advantages and disadvantages of rural residency, the participants have provided
deep insight into how rural residency hinders or facilitates a safe ‘good death’ by
creating safe spaces.
Early in the disease trajectory, creating a safe place of care (PoC) was the focus. While
patients remained well enough to travel, the tertiary hospitals and medical specialists
contributed to creating a safe space for most participants. The perceived inexperience of
local GPs in managing complex issues, such as complex pain, was often negated by
consulting with the ‘not-so-far away’ Canberra specialists. However, some patients and
their FCGs (e.g. Cheryl and Pat) did not feel safe in the city hospitals because they were
unfamiliar spaces both in terms of location, environment and personnel. Despite the
issues surrounding formal handover of care, and providing patients could return home
as quickly as possible, most patients were satisfied with alternating their PoC between
home and the most appropriate facility to provide the level of care required. Patients
were mostly willing to accept admissions to either the local or out-of-region hospitals.
Patients and FCGs accepted that rural hospitals could not provide radiation therapy,
management of acute oncology conditions such as febrile neutropenia, intensive care, or
specialist surgical procedures and interventions. While patients could travel, distances
were not a major obstacle to creating and maintaining a safe space.
However, once patients became too ill to travel, or entered the terminal phase, the issues
surrounding travel distances intensified. While a medical reason to see oncologists and
respiratory physicians no longer existed, for some participants this lack of contact with
specialists created a degree of insecurity. For some patients, this deficit in care was
adequately replaced by the local health professionals. However, Elaine and Andrew did
208
not trust the local HCPs to manage her complex pain, and without a local specialist
inpatient hospice, they made the decision to not only leave home, but to move out-of-
region so Elaine could die in a safe place:
[The out-of-region hospice] turned out to be the best choice for us. In slightly
different circumstances it would have been far better to have her at home. But
really her pain management was beyond my ability to keep track of and keep up
to scratch…I would have stayed home if I thought that I could manage her pain.
But things were changing, and I didn’t have that confidence…And the local
people didn’t have a clue. (Andrew, husband of Elaine)
Despite the strong community support and beautiful environment of home, for most of
the participants, the safe haven of home became unsafe. Ivan’s home remained safe
until the end, partly due to the support of his family, friends and HCPs, and the
resilience of his farming family. But, the major determining factor was possibly his
disease trajectory. Ivan was able to ‘live’ and participate in life until a few days before
he died. His episode of acute deterioration and restlessness lasted only a few short
hours. In contrast, Marianne had distressing symptoms for nearly three days. Her home
became an unsafe space; however, her husband perceived the local hospital to be more
‘unsafe’ than home, and so she remained at home. Had their GP friend not visited and
supplied the appropriate medication, Marianne may well have died in hospital.
Six of the patients moved from home once home became an unsafe space. Home
became unsafe because of symptoms or carer burden. The local rural hospital became a
safer space. The lack of all-hours formal support may have been a contributing factor in
some circumstances; however, the decision for George and Leanne to leave home was
made well before home became unsafe. While there were perceived deficiencies in
skills and expertise of the rural hospital HCPs, the local hospital quickly became a safe
space for those admitted for end-of-life care. This was no doubt facilitated by the value
placed on rural ‘community.’ The two rural hospitals were considered to be community
hospitals where patients received personalised care, where patients and their families
were known, and hospital personnel were known members of the wider community.
209
While rural residents are less likely to die at home than urban residents, and my
participants living on farms were less likely to die at home than those living in town,
safe spaces were created at home. My sample of rural residents illustrated that rural
homes, and especially farm homes, can quickly become unsafe spaces. While not
necessarily generalisable, it is possible that rural homes are more likely than urban
homes, to become unsafe due to the challenges of providing suitable and sustained
services and resources into rural areas. However, when home became unsafe, the two
local rural hospitals were more likely than the large tertiary hospital to be perceived as a
safer alternative to home.
The previous chapter concluded that PoD was not the main determinant of the ‘good
death’. Marianne’s husband and community nurse did not consider her death at home a
‘good’ death. While remaining home contributed to the distressing experience, the
inability to overcome the challenges of rural residency was possibly the greatest factor.
It is impossible to know the outcome had Marianne lived in town, but specialist
palliative care home support would most likely have been available at all hours had she
lived in Canberra, or any other large urban area.
I have previously indicated the elements necessary for a ‘safe good death’ (see Figure
7.1 in Chapter Seven). For some participants, like Ivan, the presence of family was
strengthened by rural residency as his son lived on the property. For others, like Gwen
and Barbara, their families had moved from the rural town and had to travel many
hundreds of kilometres to be present at the time of death. For all participants the support
of community was powerful throughout their illness. All commented that the level of
community support would unlikely have been available in a city. ‘Being known’ and
‘knowing others’ was a unique feature of rural communities, which for most patients
helped facilitate a ‘good death.’ The degree of symptom control was variable amongst
the participating patients. Those with complex pain issues were the least satisfied with
the level of local expertise. All the participants spoke of the importance of receiving
personalised care and maintaining identity. Rural residency strongly facilitated this with
the generational farmers having the strongest connection to the region. Due to the
limited choice in HCPs, resources and inpatient facilities control over many aspects of
care was limited; however, most patients and FCGs indicated they did actively
participate in decision making surrounding patient care. Where circumstances were
210
outside their control, for example Gwen who developed an unexpected hemiparesis,
patients and their FCGs often took a pragmatic approach with most eventually accepting
the change in plans.
8.6 Chapter summary
This chapter has explored the aspects of rural residency that either facilitated or
hindered a ‘good’ rural death, and determined the patients’ actual PoD. Each of the
participants had different priorities, and held different perspectives. What one
participant identified as a barrier, another considered it a challenge that could be
overcome. Some of the challenges identified included distance and the need to travel,
the lack of local specialists, limited experience and expertise of the local HCPs, and the
lack of allied health services. The challenge of travel was not just for the patient
attending medical appointments. Distance also resulted in limited availability of formal
home support services, and was an additional cost for out-of-town families wanting to
visit and be present at the time of death. Despite the challenges, there were many
positive features to rural residency that facilitated the creation of safe spaces, including
lifestyle, community, personalised care, being known, and the convenience of small
towns. While rural residency helped create home as a safe PoC in the early stages of the
disease trajectory, these same features rendered home an unsafe PoD. Where
participants once relished their isolation and privacy on the farm, this isolation often
worked against them when their care needs escalated. The rural culture of community
and ‘being known’ facilitated the adoption of hospital as a safe space. While most
patients wished they could die at home, when this was not possible, most readily
changed their preference for PoD from home to hospital.
This is the final chapter reporting the results of my research study.
211
PART 4: DISCUSSION AND IMPLICATIONS
212
213
CHAPTER NINE: DISCUSSION AND IMPLICATIONS
Introduction
The primary objective of this research study was to explore the concept of the good death
drawing on the experiences and perspectives of rural patients with a life limiting illness,
and their family caregivers (FCGs). The secondary objectives were to explore the influence
of place of death (PoD) and rural residency on the concept of the good death, and to
explore the meaning behind the PoD preferences and decision-making process.
One measure of the success of palliative care is to enable people to die in their preferred
place. The literature reports ‘home’ to be the preferred place for most (Gomes et al. 2013),
thus implying, a ‘home death’ is central to achieving a ‘good death’. Consistent with the
international findings, (Broad et al., 2013; Cohen et al., 2015), most Snowy Monaro
residents, including the interviewed patients, regardless of cause of death, died at a place
other than home, with most deaths occurring in an institution, such as hospitals and RACFs.
When ‘home’ was reclassified as ‘usual place of residence’ (UPoR), in acknowledgement
of the fact that RACFs are often considered ‘home’ by many long-term, permanent
residents (Cartwright and Kayser-Jones, 2003), most deaths still occurred in an institution.
Therefore, if the standard that a home death is central to achieving a ‘good death’, one
could conclude that most residents in the Snowy Monaro, including those who died of a
condition amenable to palliative care, did not die a ‘good death’. However, this stance was
contested by the interviewed participants.
The popular equation of the ‘good death’ with the at-home death has some limitations.
Firstly, data on preferences for PoD have typically been collected using surveys, mostly
completed by a healthy population, with limited data reporting the preferences of patients
with a terminal illness, and their FCGs. Where patients and FCGs have participated in
studies, cancer patients tend to be over-represented. The 42 in-depth semi-structured
interviews, in this research study, have provided opportunity for 11 rural patients, with
cancer and non-cancer diagnoses, and a life expectancy of less than six months, 18 FCGs
and six local HCPs, to describe their experiences and perspectives on what constitutes a
‘good death’, and the meaning of PoD within one rural setting. These perspectives have
214
provided texture and meaning to the actual PoD data, collected through the population
death data survey.
A second criticism of the current literature is that PoD data have been derived from a single
point in time. They do not capture the many patients who received their care at home, only
to transfer to an institution for the very last hours or days before death. By conducting
longitudinal interviews, with a sample of Snowy Monaro patients and their FCGs, my
research findings support those of Agar et al. (2008), and strengthen the argument that PoC
and PoD are not necessarily the same, and that preferences change along the disease
trajectory. Finally, by conducting this research in a rural setting, rural voices, often unheard
within urban focused research, have been placed centre stage. Gott et al. (2013) have
recently challenged the popular belief that end-of-life admissions to, and deaths in hospitals
are ‘inappropriate’, ‘potentially avoidable’ or even an indicator of sub-optimal EoLC (Earle
et al., 2003). Further research by Gott’s team in New Zealand found that being admitted to
an acute city hospital has identified benefits for non-Maori patients with palliative care
needs (J Robinson et al., 2012, 2017). This view is consistent with the findings in this
thesis, that most rural patients, and their FCGs and local HCPs, considered a death in the
local rural hospital appropriate, and at times necessary and planned. Depending on their
clinical situation, most participants in this research study considered the two rural hospitals
an appropriate place to die, and a facilitator of a ‘good death’.
In this chapter I discuss the findings of the death data survey by placing the findings into
the context of existing literature. This is followed by a discussion of the themes identified
in the in-depth interviews, meaning of place (section 9.2), the ‘good death’ (section 9.3),
and rural residency (section 9.4). Limitations to the study are discussed in section 9.5. The
importance of researcher reflexivity, outlined in Chapter Four, is discussed further in
Section 9.6. Implications for policy and recommendations for further research are discussed
in section 9.7.
9.1 Place of death for Snowy Monaro residents
Between 1st February 2015 and 31st May 2016 there were 224 deaths in the Snowy
Monaro region, of which 190 (85%) were due to ‘natural’ causes. Of these, 138 (73%)
were due to one of the conditions considered to be amenable to palliative care. Cancer
215
deaths are often over-represented in rural palliative care PoD studies. The data collected
in this research study describe PoD for both cancer and non-cancer deaths. Consistent
with the current international whole population literature (Broad et al., 2012), the study
data highlight the significant reliance on hospitals and RACFs for actual PoD.
In this current study, the overall rate of hospital deaths was 43.5%, RACFs 29.5%, and
private residence 19%. In 2012, Broad et al. published an international paper reporting
on the PoD for 16.2 million decedents, from 45 populations (including, but not
identified, rural populations). The proportion of hospital deaths, in my research study, is
consistent with those findings (median 45%, range 40-60%); however, the proportion of
deaths in RACFs in my study is higher than the median reported by Broad et al. (median
12%, range 0-20%). My results are consistent with some individual countries identified
in their findings: Norway (44% of all deaths occurred in a RACF), New Zealand (31%),
and Australia (26%). With no inpatient hospice in the Snowy Monaro, it is not
uncommon for patients who are unable to return home from hospital to be transferred to
a RACF for EoLC, as illustrated by the cases of Gwen, Bruce and Cheryl. Rural
communities are also aging at a faster rate than urban areas, mainly as a result of
internal migration with younger family members moving to the city for education and
work (Aged and Community Services, 2004). As a result, rural residents are often less
likely to have family to provide EoLC at home. This places a greater reliance on
RACFs, especially for those aged 65 years and older. Three Australian rural studies
(Burns, 2015; Crawford, 2000; Howat et al., 2007) and one Canadian rural study
(Jayaraman, 2013), reported home death rates of 18%, 26.7%, 19% and 20.2%,
respectively. These are similar to the Snowy Monaro rate of 19%.
Over half of the residents in this study, who died of a palliative condition, died in
hospital (52%). The high rate of deaths in hospital, especially in the developed world,
has been attributed to the medicalisation and institutionalisation of death (Aries, 1975;
Kellehear, 2008). However, as illustrated by the interviewees in this thesis, this is only
part of the reason. People are living longer, often with chronic diseases and debilities.
Society itself is changing. Previously, mothers stayed home to care for the children, and
often their elderly relatives. Today, both parents are often in the workforce. Similar to
urban families, rural families are often fragmented, due to family breakdown, or adult
children moving away for work. Many adult children in this study were working or
216
living away, some interstate or overseas, and while some could offer intermittent
support, most were unavailable full-time. While Ivan’s and Kevin’s daughters could
return home to help care for their respective fathers, this would not have been
sustainable long term. Bruce’s daughter lived nearby, but was under tremendous strain
juggling a small business and care.
There are large differences between countries in the reported rates of home deaths for
cancer and non-cancer decedents (Cohen et al., 2015). While Cohen et al. (ibid)
reported residents who died of cancer in England, New Zealand, Canada, and the USA,
were more likely to die at home than those who died of a non-cancer cause, this is not
consistent with my findings. In the Snowy Monaro, those who died of cancer were more
likely to die in hospital (65%) than those of a non-cancer cause of death (40%). Despite
cancer patients having a more predictable disease trajectory (Murray et al., 2005), the
higher rate of hospital deaths could again be due to the ‘rural’ effect, and limited formal
in-home support, especially after hours.
In the Snowy Monaro, the proportion of home deaths for cancer patients was 14%, and
for non-cancer conditions amenable to palliative care 11%, where ‘home’ implied a
private residence. When the term ‘usual place of residence’ was used to refer to a
private residence or permanent RACF, as suggested by McNamara and Rosenwax
(2007), the rates for cancer were 26%, and non-cancer 49%. Within the current rural
literature, there is great variation in the rate of home deaths across studies, with more
traditional regions reporting higher rates of home deaths. Studies in southern Italy
(Constantini et al., 2000), Spain (Catalan-Fernandez et al., 1991), and Mexico (Castillo-
Guzman et al., 2013) reported home death rates, for residents with cancer, to be greater
than 50%; however, in these studies, there was no RACF category, with deaths in
RACF either absent or contained in the hospital category. One Swedish rural cancer
study (Axelsson & Christensen, 1990) reported a home death rate of 12%, with the
remaining studies clustered around 25%. In the Snowy Monaro study, of the 65
permanent residents in RAC, 78.5% died at the facility. This rate was 84% for those
with dementia. Therefore, care must be taken when interpreting the rate of ‘home’
deaths, as most permanent residents of RAC in my study, died at their facility, that is,
they died in their UPoR, and for many that was ‘home’.
217
The reason for reporting the rate of deaths at home should determine which terminology
is used. If the purpose is to determine the quality and effectiveness of the overall EoLC
provided, then UPoR is a more appropriate term, as this rate identifies all who died ‘in
place’ and not just those who died in a private residence. However, if the purpose is to
inform service providers of the needs within the community for resources to assist
people to remain, and be cared for, in their private residence then ‘home’ is the
appropriate term. When RAC permanent residents are excluded from the ‘home’
category, this has the potential to skew the data by falsely increasing the rate of deaths
within institutions. While the interviewed RACF residents considered the facility to be
‘home’ this may not be true for all residents. With length of stay in RACFs decreasing,
as people remain home longer and until all other options have been exhausted, and some
older patients (over the age of 65 years) transferring from acute hospitals to RACFs for
end-of-life care, Phillips and Currow (2017) have questioned the validity of calling
RACFs ‘home’. Further research is encouraged to explore the meaning of ‘home’ for a
cross-section of permanent rural RAC residents and their FCGs.
The findings in this thesis demonstrated that cause of death is associated with PoD,
especially for those who died in their UPoR. Residents who died of cancer were less
likely to die in RAC than those who died of a non-cancer cause. Compared to patients
with cancer, residents with dementia were more likely to live in a RACF, and therefore
die in their UPoR. Access to health–system facilities, often lacking in more rural and
remote regions, was a key factor in determining POD, with residents living on farms or
in small towns, with limited or no medical services, being less likely to die in their
UPoR than those living in towns with general practitioners and rural hospitals.
Inconsistent with the literature, age, sex, and marital status were not predictors of PoD
within the Snowy Monaro; however, data on marital status were missing for 32% of
deaths.
The results of the PoD analysis raise several questions. Of the residents who died of a
condition amenable to palliative care, did the 17 who died at home have a ‘good death’?
Did the 72 who died in hospital have a ‘bad death’? Did those who died in a RACF die
at ‘home’ or in an institution? These data do no more than provide a snapshot of current
practice. While this is important, the quality of care provided and received, the
congruence between actual and preferred PoD, the changes in preferences along the
218
disease trajectory, the meaning behind the final place, cannot be ascertained from the
statistical data.
9.2 Meaning of place
All 15 patients, identified through the interviews, had a preferred PoD, with six dying in
their UPoR (four at their private residence, and two in a RACF). Consistent with the
findings of Agar et al. (2008), place of care (PoC) and PoD were not necessarily the
same. There was good congruence between the preference of the patient and their FCG
(Davies et al., 2016). Most of the participating patients, supported by their FCGs,
accepted the need to alternate between home and hospital for management of
symptoms, such as blood transfusions, providing the length of stay was kept to the
minimum. Initially, some patients were prepared to travel out of region; however, once
the patient became too unwell, and travelling too burdensome, the local rural hospital
was considered appropriate for short admissions. Providing people could remain home,
for as long as was possible, once home became unsafe, their preferences regarding PoC
and PoD changed, and home as the actual PoD became less important than the literature
suggests. This illustrates the inadequacy of relying on data obtained from surveys of
healthy populations, when many do not have first-hand experience of dying, or caring
for someone at home at EoL. In addition, when studies have reported only a one-time
preferred PoD, the full story is left untold.
Participants associated home with identity, self-worth, connection, memories, family,
social contact, control, normality, and familiarity. More importantly, home was
determined to be a safe place (Collier et al., 2015). In 2005, Morris and Thomas linked
safety with the ‘right place’ to die. Over time, the connection of safety with place, has
been overshadowed by concepts such as autonomy, choice, avoiding hospitalisation,
improving health budgets, and reducing hospital costs. Researchers have begun to
explore the often invisible or uncountable cost-shifting from hospital to the FCG, when
EoLC is provided at home (Gott et al., 2015; May, 2017; Rowland et al., 2017). Such
studies suggest that there are economic advantages to health care institutions, and
economical and personal costs for FCGs, in delivering home-based palliative care.
219
Interviews with families and the dying patients in this thesis indicated that the preferred
place is the ‘safe place’, regardless of its physical location. Once circumstances
rendered home unsafe, the participants in this study either chose or were encouraged by
their FCGs or HCPs, to leave home. Five transitioned to one of the local rural hospitals,
three to a RACF, and one to the out-of-region hospice. The new site of care was mostly
accepted, providing the patient and their FCGs perceived the new place to be a safe
place. For most patients and FCGs, the move from home did not have a negative effect
on the quality of dying. For Marianne, (Chapter Eight), home came to be seen as unsafe
by the family and the community nurse. However, hospital was considered to be more
unsafe by the family, and so she died at home, but with negative consequences for her
husband.
For the patients and their FCGs, home became unsafe for various reasons. The most
frequently cited reasons for leaving home were symptom control and carer distress. For
most FCGs, being a carer was a privilege and the final act of kindness for their family
member; however, some FCGs were overwhelmed by the responsibility, uncertainty,
and feelings of inadequacy (Seal et al., 2015). In contrast to the HCPs interviewed in
one New Zealand urban hospital study (Gott et al., 2013), who considered admission to
hospital for carer inability to cope inappropriate, the HCPs in this current study
regarded admission to the local district hospital as appropriate. Ivan’s GP, expressing a
commonly held view, suggested that patients and their FCGs often suffer unnecessary
stress by remaining home, and that hospital admissions should occur earlier and more
often. Again, this is possibly the ‘rural’ effect, as full time medical and nursing support
at home is not available, and the hospital is a substitute for in-patient hospice. The rural
effect has the potential to render rural homes, especially farms and outliers, more
susceptible to become unsafe, earlier and more frequently than most urban homes.
Safety is central to the delivery of healthcare, regardless of place of care. The National
Health Performance Committee defines safety in the health care system as ‘the
avoidance or reduction to acceptable limits of actual or potential harm from health care
management or the environment in which health care is delivered’ (AIHW, 2017). In
August 2012, the Australian Safety and Quality Goals for Health Care were developed,
mandating that ‘people receive health care without experiencing preventable harm’
(Australian Commission on Safety and Quality in Health Care, 2012). Within the
220
context of healthcare organisations, Vincent et al. (2014, p. 670) suggest safety is partly
achieved by being alert to perturbations [failures in the functioning of the system] and
responding rapidly to keep things on tract’. Safety in hospitals and other healthcare
institutions is often equated with compliance and assurance (Vincent et al., 2014, p.
671).
The interviewees in this thesis described safety in different terms. For them, safety
implied protection for the patient, and ensuring a psychological sense of security during
an unstable and complex physical process. Most participants in this thesis indicated the
local hospitals were an appropriate safe PoD when home became unsafe. If hospitals are
to be safe for the terminally ill it is assumed that policies are in place to maintain safety.
However, creating a safe place is not just following organisational policies and clinical
procedures. Safety is subjective and individualised (J Robinson et al., 2017). Marianne
and Kevin, who had complex pain symptoms, did not feel safe in hospital, because they
and their FCGs lost control of their medications. While hospital policies surrounding
drug administration ensure protection for patients and clinical staff from medication
errors, the policies also mean pain medications are often not available immediately, or
as quickly as families can respond at home. While organisational policies ensure clinical
systems and processes are reliable, the implementation is dependent on ‘the capacity of
staff to follow safety critical procedures…monitor safety…anticipate and be prepared
for problems and threats to safety… detect, analyse, integrate, respond and improve
safety’ (Vincent et al., 2014, p. 671). For some participants, including the HCPs, the
ability of the local hospitals to detect, analyse and respond, made hospital safer than
home. This is possibly more significant in rural areas, where there is often no full-time
home healthcare and medical support to respond to changes in patients’ conditions.
Rural hospitals could be perceived as less safe than urban hospitals, due to the lack of
specialist palliative care and expertise; however, this was not the perception of the
research participants. The familiarity of, and personal attention provided by, the rural
hospitals, and the benefit of being known, created a safe place within the rural hospitals.
However, this experience was not universal. When the local hospital was perceived as
unsafe, Marianne remained home, possibly by default rather than choice, while George
and his sister, Betty, suffered in hospital, and Elaine left her community to die in the
out-of-region hospice. For patients and their FCGs, safety is not simply following
procedures and policies. Safety is aligned with trust. Trust is based on prior knowledge
221
of the hospital, and dependent on the moment-to-moment experience of care, not only
for the patient but also their FCG. Safety is ‘conveyed by the manner of staff, empathy
and compassion’ (Vincent et al., 2014, pp. 673-4).
Previous studies have associated acute hospitals with ‘aggressive and futile treatments’,
and over medicalisation (Gott et al., 2013). This was not how the participants perceived
the two local rural hospitals. Gott reports that acute hospital renal nurses felt the
hospital was appropriate for EoLC for their renal patients who, over time had developed
relationships with the staff, and become familiar with the hospital. While the
participants in my study identified the tertiary hospital as appropriate for specialised,
aggressive, and hopefully life prolonging treatments, such as radiotherapy for
hemiparesis due to new brain metastases, febrile neutropenia secondary to
chemotherapy, intensive care for previously undiagnosed cerebral lymphoma, it was
never a familiar or homely place. Despite the presence of specialists and expertise, the
tertiary hospital was frequently perceived as unsafe. For most participants, the local
hospital was, or became, a safe place, and while some patients chose to remain home,
all participants, except Marianne’s husband and Elaine, believed the local hospital to be
an appropriate place if the circumstances required an admission.
In the same way, RACFs need to be safe for the terminally ill. Gott et al. (2013)
reported families believed a transfer to hospital in cases of acute deterioration showed
the staff were providing good care. This was not the perception of the patients, FCGs
and director-of-nursing (DON) I interviewed. The participants in the current study
thought the RACF could, and did, provide appropriate EoLC, and that the hospital could
not have provided different or better care. However, my views have been informed by
only three patients, four FCGs and one DON, and none had experienced acute
deterioration or significant and uncontrolled symptoms. However, all the RAC
participants identified the lack of staff, especially registered nurses overnight, as a
negative influence on care, and therefore safety (Wetle, 2005). Gwen’s daughter raised
the concern of lack of staff experienced in monitoring pain and other symptoms at EoL.
Fortunately, Gwen did not have significant pain before she became unarousable and
unable to ask for pain relief. One can only assume this continued until her death, and
hope that medication was given; however, the literature reports the frequent under-
estimation of pain by RACF staff due to lack of expertise (N Johnston et al., 2016).
222
Despite this short-coming, Gwen’s daughter believed the right decision was made for
Gwen to remain in the facility, and that she died a ‘good enough’ death.
For preferences to be meaningful, there needs to be genuine choice. While the literature
reports inpatient hospices are considered by some patients and FCGs as an appropriate
alternative to home (Thomas et al., 2004), in this setting there is no local inpatient unit.
This potentially accounts for the high percentage of hospital deaths within the Snowy
Monaro region. The lack of inpatient hospice possibly had minimal effect on the rate of
deaths in RACFs, as dementia was the most frequent cause of death amongst residents,
with previous studies reporting most deaths due to dementia are likely to occur in an
aged care facility or acute hospital. (Badrakalimuthu & Barclay, 2014). However, for
Gwen, who transferred from hospital and died in a RACF, a local inpatient hospice may
have been a more acceptable option. Only two interviewed patients, and one identified
through the bereavement interviews, considered the out-of-region hospice a genuine
alternative to home. Only one participating patient died in the out-of-region inpatient
hospice. The hospice was recorded as the PoD for four decedents in the population
death data survey; however, it is impossible to determine if this out-of-region place was
by choice or circumstance. A number of interviewed participants discussed the hospice,
but it was dismissed as an option due to its location out-of-area. The hospice was an
unfamiliar place for the remaining participants, and so did not feature in the interviews.
Morris and Thomas (2005, p. 21) suggest PoD is not a matter of individual choice, ‘but
rather embedded in pre-existing relationships with place and other people’. Home can
quickly become unsafe, especially out of hours and the further one lives from HCPs and
the hospital. Home is often unsafe for patients with a terminal illness, living on their
own, in advanced stages of disease, or if symptoms become uncontrolled. Safety is
often associated with rapid access to appropriate symptom management.
Congruence between preferred and actual PoD is an appropriate measure of quality,
providing the recording of preferences is kept up-to-date. Except for possibly George,
who in his delirium repeatedly requested to return home, all the participants in this
study died in their final preferred place, as determined by them towards the end of their
terminal illness. This place was mostly not their initial preference. McCall and Rice
(2005) suggest that wishes and preferences are not the same, and this notion fits well
223
with the impression gained through the interviews. For those participants who died
away from home, home most likely remained their ‘heart preference’, while the new
place became their ‘head preference’. The reality of the circumstances at home dictated
a move from home.
Changes in preferences raises concern over the current practice of health data collection.
In Australia, systems are in place to record PoD, with these data used to measure the
quality of palliative care health services, and to compare services against benchmarks.
PoD data are usually dichotomised into died ‘in preferred PoD/ not in preferred PoD’.
This recording system does not allow for any explanation for changes in preferences or
changes in location. If medical records indicate a person’s preferred PoD was home, and
they died in a hospital or hospice, this is considered by some to be a ‘failure’ in care,
despite the patient remaining home until the very last hours or days. The literature
reports some FCGs suffer adverse effects as a result of not honouring requests to die at
home (Seal et al., 2015); however, while some FCGs were ambivalent about the move,
none of those interviewed in the bereavement period indicated they had failed in their
provision of care because their family member transferred to hospital. For three of the
participants who died at home, the PoD could have so easily have been different. Ryan
deteriorated suddenly with acute confusion and agitation. This was distressing for his
family, and would have been unmanageable at home without immediate formal support.
It was only by chance that I lived across the road and was immediately available to
administer medication. Ivan also developed acute confusion and agitation. Again, it was
fortuitous his GP was a friend, and available, on his day-off, to visit the farm. Marianne
died at home, only because their friend was a GP and provided medical and emotional
support overnight. Under normal circumstances, the ambulance would have been called
for all three patients, and they would have been transported and admitted to the local
hospital for their last few hours of life. However, with nursing and medical support at
home, these brief episodes of terminal restlessness were managed at home. Despite
formal support, had the symptoms been prolonged, home support would not have been
sufficient, and the professional advice would have been to move from home, which had
become unsafe, to the safer option of the hospital.
224
9.3 The rural ‘good death’
A ‘good death’ was difficult to define for the research participants. Marianne’s story,
(Chapter Eight), demonstrates that the concept is subjective, and dependent on personal
values, perspectives and priorities. For some participants, the ‘good death’ referred to
the dying process, highlighting the importance of living life, while maintaining
normality and quality of life. For others, the ‘good death’ referred to the moment of
death, or even the after-life. The factors associated with a ‘good death’, among the
interviewees, were not uniquely rural and were consistent with those identified in the
literature review. They included adequate symptom control, presence of family, support
of health care professionals and informal carers, a sense of life completion and
acceptance of death, all while maintaining a sense of autonomy and control. Dying at
home was initially considered an important factor for all the participants; however, over
the course of their illness, home became less important for most (Davies et al., 2016).
This contradicts the popular contemporary Western view that prioritises home as the
preferred PoD, and associates home with the ‘good death’. It is also unclear if the 100%
initial preference for ‘home’ is a genuine preference, or a default due to the lack of
alternative locations, such as inpatient hospice.
Congruence between final preferred and actual places of care and death was considered
important to most patients and FCGs, and to a lesser degree by the health care
professionals. Most participating patients had an idealised, or even romanticised, view
of the ‘good death’; however, the realities of the ‘good death’ were, for this study,
narratively reconstructed by the FCGs after the death of their family member. For
FCGs, the ‘good death’ in reality was sometimes at odds with the idealised notion of a
‘good death’ held by the patient.
In rural settings, relationships between place and one’s self are often stronger than for
urban residents, so one would expect that rural people would view dying at home as a
major feature of the ‘good death’. However, home as the actual PoD was not as
important in reconstructing the ‘good death’ by most FCGs after death. Dying a safe
death was central to the ‘good death’. The rural participants, in this study, placed a
higher priority on safety than home, with most participants willing to compromise and
settle for a ‘good enough death’ (McNamara, 2004), to ensure safety.
225
9.4 Influence of rural residency on the ‘good death’ and place of death
In a recent narrative synthesis of rural and remote studies, Kirby et al. (2016), reported
that reduced access and availability of services in rural and remote areas has a negative
influence on outcomes. From some perspectives, for example those of Marianne and
George, the lack of alternative PoD which forced hospital to be the default to home,
could be seen as a negative outcome. However, as previously discussed, the dependence
on the local rural hospital was for most participants, a good outcome. Rural residents
are frequently reported to be more resilient than urban residents, more accepting of their
circumstances, and more supported by community networks (Kirby et al., 2016). These
characteristics were demonstrated in the interviewed Snowy Monaro residents. While
Kirby et al. (2016) suggest the rural strengths are ‘counterbalanced by low expectations,
“making do” and a reluctance to seek help’ (p. 297), this perspective was not a major
feature of the Snowy Monaro participants, with most having high, but realistic,
expectations for the care they would receive. For most who had received treatment in
urban hospitals, the positives of the rural care outweighed the negatives.
The tyranny of distance, associated with rural residency, could be regarded as a barrier
to achieving positive outcomes; however, distance was not necessarily measured by
mileage. While commuting for treatment, or travelling to visit patients at home, were
costly for patients, FCGs and HCPs, in terms of dollars, time and energy, there were
significant benefits in travelling (Pesut et al., 2010), and the inconvenience was
accepted by most rural residents in this study. Kirby et al. (2016) reported that rural
cancer patients and FCGs had similar types of needs (information and personal
interaction with HCPs), and FCG burden, as urban residents. Most of the patients and
FCGs spoke of the personal support they received by ‘being known’ by the HCPs and
the wider community. While the Snowy Monaro region has unique community supports
(MCCR and Mary Green room), there was a sense amongst some participants that the
community could be more supportive, especially for those living out of town. The
literature describes the community as an ‘untapped resource’ (Horsfall et al., 2013).
226
9.5 Limitations
There are several limitations to this research study. Firstly, the population data were
collected over a relatively short period of time. While the data are representative of the
whole population, it is uncertain if the results would be different over an extended
collection period. Many residents in the Snowy Monaro are involved in seasonal
activities, such as farming or winter tourism; however, due to the short data collection
period any potential seasonal differences in PoD were not observed. Limitations due to
inaccessibility to official death certificates were discussed in Chapter Five. However,
with access to a variety of alternate sources, some cross-checking was possible; I
estimate that death data were collected for close to all deaths during the 16-month
collection period.
The research study’s sample size consisted of 35 participants. This is a robust sample
size which represents a wide range of residences, modes of dying, causes of death and
places of death. I was able to interview some participants repeatedly through the course
of their dying, and to triangulate these interview data with interview data from family
caregivers and health care professionals. This all helped the data to be valid, though, as
with all qualitative research, there is often a generalisability-validity trade-off. The
results may not be generalisable to all rural locations, or even those of comparable
populations However, the broader notion of safety being the dominant driver for
identification of the best place of death is likely to reflect concerns of dying people in
other regions.
While the Snowy Monaro region lacks a specialist multi-disciplinary palliative care
team, it is well resourced with health services, and is in close proximity to specialist and
tertiary services in Canberra. Therefore, the results are unlikely to reflect the
experiences of those living in remote regions. In the absence of a local specialist
multidisciplinary palliative care service and inpatient hospice, the perspectives gained in
this study may reflect the lack of familiarity of residents with the services offered by
specialist teams and inpatient hospice. However, many of the participants were aware of
palliative care services and inpatient hospices having looked after family members in
other settings.
227
The literature reports the significance Indigenous peoples place on being connected to
land, family and community, the importance of a home death, and the burdens faced
with hospitalisation (Gott et al., 2013; McGrath, 2006, 2007; Robinson et al., 2012).
Many of the Snowy Monaro Ngarigo people no longer live on-country, though they
retain their connection to land (ABC News, 2016). Very few residents identify as
Indigenous (Aboriginal and/or Torres Strait Islander people [ATSI]), with the 2016
census reporting the proportion of ATSI in the Snowy Monaro less than the NSW
average (ABS Census QuickStats 2017a). Official death records indicate if a decedent
identifies as ATSI; however, as discussed in Chapter Four, access to these records was
denied. None of the interviewed participants were of ATSI descent, and, due to strict
guidelines surrounding recruitment of ATSI persons, ethics approval prevented me from
purposefully selecting ATSI participants. Therefore, it is not possible to make any
comment or compare my findings to studies reported in the literature relating to the
experiences and perspectives of Indigenous rural residents.
9.6 Reflexivity
To enhance the validity of qualitative research, it is essential for the researcher to
employ reflexivity, reflecting on how their own experiences and beliefs may have
impacted on the collection and interpretation of data. As a Snowy Monaro resident and
practicing medical practitioner (palliative medicine specialist and GP), I potentially
brought to this study personal and professional bias. There was also the potential for
conflict of interest as doctor/researcher. From my perspective, the best place to die is
not necessarily home, but I endeavoured to step outside my pre-conceptions and to
welcome and reflect on the perspectives of patients and FCGs in their own contexts and
settings. In order to keep track and monitor my own perceptions, and to identify the
intrusiveness of pre-conceptions, I kept a personal journal throughout the interview
process. I present my reflections in the hope the reader of this thesis may have
confidence that my professional pre-conceptions have not directed the course of the
findings.
One concern of the ethics committees was that as a doctor/researcher, and the only
palliative medicine specialist in the region, patients may feel coerced into participating
and offering their perspectives. At the start of every interview I reminded the patient
228
and their FCG that I was acting as a researcher and not their doctor. I also advised that
my direction of questioning did not imply anything was, or was not, being considered or
would eventuate. Knowing some of the participants added depth to the interviews.
Participants were open in their discussions, trust was already established. From my
perspective, it was easy to be empathic as I knew, or in some cases I wrongly thought I
knew, their personal struggles and what they were going through. It never ceased to
amaze me the depth of new information I was privileged to hear; experiences and
perspectives that are too lengthy for patients and FCGs to narrate within the time frame
of a normal 15 minute clinical consultation.
Rapport was established quickly with all the participants, both known and unknown. All
the participants were open and appeared to enjoy the interviews. There was never a
sense topics were avoided or responses altered to provide me what participants thought I
might want to hear. I was told of things that went well, and things that were done
poorly. I was left with the impression that people, especially the bereaved FCGs,
appreciated the opportunity to talk, to be heard, and to ‘get things off their chest’,
especially when outcomes were not as planned. Silences were embraced, and were often
followed by something profound.
The patient participants were vulnerable and unwell, the FCGs burdened with the
responsibility of care, the HCPs were busy. All this had to be considered during the
interview process. While Kevin was happy to be interviewed, his wife and daughter did
not want Kevin tired out. They requested a copy of the questions before I interviewed
Kevin, so he could prepare his responses in his own time. I respected this request
knowing that Kevin was so unwell and easily fatigued. His was the shortest interview,
and the only one with a set agenda. Having cared for Kevin and his wife for the
previous few months meant I could dispense with the preliminary ‘chit chat’ and just
ask direct questions relevant to his perspective on the ‘good death’ and his preferred
PoD. This would have been more difficult had I not had an established rapport with
Kevin and his family.
Interviewing patients and their FCGs also had an effect on me, both personally and
professionally. During the interviews, I became aware of the participants’ expectations,
and on a few occasions, as their treating palliative care specialist, I was concerned I
229
would be unable to live up to their expectations. As Gwen re-encountered her
experience, many years ago, of caring for her mother, I knew I could not “always [be]
there, [available] to visit every day [at her] beck and call.”
Gwen’s sudden deterioration was stressful, not only for Gwen and her daughter, but also
for me as a friend and doctor. I was torn between respecting her deteriorating health,
and wanting a follow-up interview. My role as her doctor was to navigate Gwen through
the decision making and treatment options. However, this was also a significant time in
my research as Gwen’s choices regarding place of care had changed suddenly, without
any warning. It became obvious her preferred PoD at home would not be achievable.
Was it right to request an interview when she was at her most vulnerable? On the
second day, I eventually asked Gwen, and without hesitation she agreed to be
interviewed the following day, providing she remained well enough. If Gwen
deteriorated overnight I would still visit, but just as her doctor. To ensure Gwen had not
felt coerced, I sought permission from her daughter, who replied, “Yes, the distraction
might be good for her”. Gwen was now afraid of what lay ahead, and as her doctor, she
was depending on me. I didn’t want to abuse the trust. The next day when I arrived at
the hospital, Gwen was waiting for me, and eager to start the interview. It was a difficult
interview. Gwen teared up when talking about home. At one point, I switched off the
recorder for 10mins, as she spoke confidentially about her family. She told me when we
could restart. Gwen had capacity to provide consent. She knew she could decline an
interview, something she did when I later requested a fourth interview. Gatekeeping is
well documented in the literature (Bullen et al., 2014; Rainsford et al., 2015). It occurs
frequently when carers, professional and informal, deny patients opportunity to
participate in research, on the assumption they are too frail, or too unwell and therefore
need protecting from the burden of research. I knew I shouldn’t assume Gwen was too
unwell. Nevertheless, I wrestled with internal conflict regarding the appropriateness of
requesting an interview.
There were other issues I found challenging. Despite having confirmed I was
conducting the interviews as a researcher and not their doctor, at times I found it
difficult when not offering advice or comment on aspects of their management and care.
On occasions, I felt my questions forced patients and FCGs to justify their responses
and decisions. I often silently agreed with their decisions, but asked delving questions to
230
ascertain the reasoning behind their decisions. At other times, there was a fine line
between maintaining sensitivity, and ‘pushing’ participants to get to the heart of the
matter, in my search for the deep meaning and importance of home. In Chapter Four, I
discussed the conflict surrounding confidentiality and not being able to share
information gained through interviews in clinical situations and family meetings.
Before the interviews, I thought I knew my patients and colleagues well. What I learnt
through the interviewing process is that in the busyness of clinical practice, we just
don’t have the time to really get to know our patients. We don’t have time to just sit and
listen to their hopes, their fears, their experiences and perspectives. There is so much we
can learn from our patients and their families, and our colleagues, if only we had the
time. My experience illustrates the importance, of researchers not relying on surveys
stating hypothetical wishes, but the need for more qualitative studies to gain the
perspectives from those living the dying process.
9.7 Implications for policy and practice
The findings of this study illustrate that rural residents do hold unique perspectives on
EoLC, which are at times, contrary to the urban view. It is not simply the physical
challenges faced, such as travel distances, access to health services, and reduced
resources that makes the difference. The uniqueness of ‘rural’ is intangible, often
invisible, and associated with a strong sense of community and belonging. In the same
way, safety is not just about policies and standards. While urban acute hospitals may be
inappropriate, although there is recent evidence suggesting this may not be the case (J
Robinson et al., 2012), the rural hospital is more than an urbanised health service. Rural
hospitals are staffed by community members; they are a familiar place, and hold a
significant position within rural communities. There is an expectation rural hospitals
will, and currently do, provide good EoLC for the community.
While continuing to encourage governments and health services to expand resources to
support people to die at home, if that is their preference, it cannot be ignored that rural
hospitals play a different role to city hospitals; they are often a substitute for in-patient
hospice. As such, rural hospitals need to be a safe place. While the participants thought
the two local hospitals mostly did a good job, there is a need for improvement. In the
231
Snowy Monaro, the highly skilled palliative care nurse provides a consultancy service
in the community and hospitals. However, she is not available for every shift, and not
routinely available afterhours. If the rural hospital is to provide excellent EoLC for all,
then every shift should have access to palliative care support. Since I commenced my
research study, new resources have been implemented; there is now an all-hours
specialist palliative care phone support service. However, this is not the same as having
expertise on the ground. One way of addressing this gap would be to find palliative care
champions amongst the existing nursing staff. These staff members could be upskilled,
and rostered on most, if not all, shifts.
One criticism of the hospital was its strict rules governing administration of
breakthrough medications. Even when medications are charted, there are often delays in
receiving medications. If breakthrough medication is not provided immediately, pain
can easily escalate. Patients and FCGs are often familiar with administering medications
at home. Providing the patient, and their FCG, are competent, one recommendation
would be for the nurse to draw up, for example, two breakthrough doses, for either self-
administration, or in a locked bedside drawer for immediate use by the nurse, when
required.
More needs to be done to improve the EoLC in RACFs. Again, most participating
Snowy Monaro aged care residents, believed their care was good, with one regarding it
excellent. However, we need to ensure these are safe places to receive end-of-life care.
The main criticisms related to under staffing, and inexperience in managing symptoms,
especially in the unconscious patient. As in the hospital setting, palliative care
champions need to be identified, encouraged and upskilled.
While this study has added to the international literature, by exploring the ‘good death’ and
PoD from the perspectives of rural patients and their FCGs, in life-limiting illness, further
research is required. From the experiences and perspectives obtained from the interviewees,
a ‘good death’ is a safe death. However, ‘safety’ is subjective and dependent on one’s
perspective and priorities. At times, FCGs priorities and perspectives were at odds with
those of the patient. HCPs priorities were often focused on symptom management and the
desire to relieve the FCG of the burden and responsibility of providing end-of-life care. The
‘good death’ was often reconstructed by the FGC after the death of their family member.
232
Place of death was a contributing factor to a ‘good death’; it was not the main factor, and a
‘home’ death was not essential for a ‘good death’. Rural areas are unique, and it may be
that the perspectives obtained in this study do not represent the perspectives of other rural
or remote regions. National and international collaborative work is needed.
While there are features of rural residency, such as travel distances and limited availability
of all-hours health services, that restrict the provision of end-of-life care at home, there are
compensating features. Rural hospitals are an integral part of the rural community, and as
such have the potential to be a safe place, and therefore an appropriate and acceptable place
for rural residents to experience a ‘good death’. Therefore, further research is required to
determine what makes a ‘place of dying’ safe from the perspectives of patients and FCGs,
and what are the elements of a safe death. If care is to be patient-centred then processes
need to be in place to ensure that all places providing EoLC are able to access and deliver
this mixture of elements to ensure safety. The participants have illustrated that safety is not
just about policies and quality assurance. Therefore ways to develop and maintain trust
between the institution, the health care professionals, the FCG, and the dying patient need
to be explored. The foundations of good palliative care need to be in place well before the
person finds out that death is approaching.
As I complete this thesis, I argue that a death at home should not be a gold standard, or
a measure of the success, of quality palliative care. The preferred place is the safest
place, and that can’t be decided before time. Attitudes and words need to change. Policy
makers need to stop implying that a ‘good death’ is a death at home. Patients and their
FCGs need to have genuine options. Home needs to cease being the default because the
hospital is not safe. The task for policy makers, administrators, institutions, healthcare
professionals, communities, and family caregivers, is to ensure all places for dying are
safe.
233
REFERENCES
Aabom, B., Kragstrup, J., Vondeling H., Bakketeig, L.S., & Stovring, H. (2005).
Population-based study of place of death of patients with cancer: implications
for GPs. British Journal of General Practice, 55(518), 684-689.
Addington-Hall, J., & McPherson, C. (2001). After-death interviews with
surrogates/bereaved family members: some issues of validity. Journal of Pain and
Symptom Management, 22(3), 784-790.
Agar, M., Currow, D.C., Shelby-James, T.M., Plummer, J., Sanderson, C., &
Abernethy, A.P. (2008). Preferences for place of care and place of death in palliative
care: are these different questions? Palliative Medicine, 22(7), 787-795.
Agar, M., Ko, D.N., Sheehan, C., Chapman, M., & Currow, D.C. (2013). Informed
consent in palliative care clinical trials: challenging but possible. Journal of
Palliative Medicine, 16(5), 485-491.
Aged and Community Services Australia and the National Rural Health Alliance.
(2004). Older people and aged care in rural, regional and remote Australia.
Discussion paper; viewed 25 June 2017
<http://www.pc.gov.au/inquiries/completed/ageing/submissions/national_rural_healt
h_alliance_inc./sub012.pdf>.
Aries, P. (1975). Western attitudes towards death from the Middles Ages to the present.
(Translated by PM Ranum). John Hopkins University Press, Baltimore.
Arksey, H., & O’Malley, L. (2005). Scoping studies: towards a methodological
framework. International Journal of Social Research Methodology, 8(1), 19-32.
Arnaert, A., Seller, R., & Wainwright, M. (2009). Homecare nurses’ attitudes toward
palliative care in a rural community in western Quebec. Journal of Hospice and
Palliative Nursing, 11(4), 202-208.
Australian Bureau Statistics (ABS) 2015. 3303.0- Causes of death, Australia; updated
28/09/2016, viewed 18/06/2017.
<http://www.abs.gov.au/ausstats/abs@.nsf/Lookup/by%20Subject/3303.0~2015~Ma
in%20Features~Australia's%20leading%20causes%20of%20death,%202015~3>.
Australian Bureau Statistics (ABS) 2017a ‘Census QuickStats’; viewed 27 June 2017.
<http://www.censusdata.abs.gov.au/census_services/getproduct/census/2016/quickst
at/101031014?opendocumen>
Australian Bureau of Statistics (ABS) 2017b. ‘Population clock’; viewed 18 June 2017.
<http://www.abs.gov.au/ausstats/abs@.nsf/0/1647509ef7e25faaca2568a900154b63?
opendocument>.
Australian Bureau of Statistics (ABS) 2017c. Regional stats. Updated 31st March, 2017,
viewed
05/05/2017<http://stat.abs.gov.au/itt/r.jsp?RegionSummary®ion=12050&dataset
=ABS_REGIONAL_LGA&geoconcept=REGION&datasetASGS=ABS_REGIONA
234
L_ASGS&datasetLGA=ABS_REGIONAL_LGA®ionLGA=REGION®ionA
SGS=REGION>
Australian Bureau of Statistics (ABS) 2017d. 3218.0 Regional population growth,
Australia. Data cube: Population estimates by remoteness area, 2006 to 2016.
Released 30/03/2017, viewed 18 June 2017.
<http://www.abs.gov.au/AUSSTATS/abs@.nsf/DetailsPage/3218.02015-
16?OpenDocument>
Australian Government Treasury 2010. Australia to 2050: future challenges. The 2010
intergenerational report overview; viewed 02 April 2017.
https://archive.treasury.gov.au/igr/igr2010/Overview/pdf/IGR_2010_Overview.pdf
Australian Health Ministers Advisory Council 2011. National framework for advance
care directives
http://www.coaghealthcouncil.gov.au/Portals/0/A%20National%20Framework%20fo
r%20Advance%20Care%20Directives_September%202011.pdf ; viewed 21 January
2017.
Australian Institute of Health and Welfare (AIWH) 2004. Rural, regional and remote
health: a guide to remoteness classifications. AIHW cat. no. PHE 53. Canberra;
viewed 11 September 2016.
<http://www.aihw.gov.au/WorkArea/DownloadAsset.aspx?id=6442459567>.
Australian Institute of Health and Welfare (AIWH) 2015. ‘Growing Older’, Section 6.3.
Australia’s welfare 2015, viewed 02 April 2017. http://www.aihw.gov.au/australias-
welfare/2015/growing-older/#t4
Australian and New Zealand Society of Palliative Medicine 2014. ANZSPM position
statement on quality end-of-life care- part 1: essential elements for quality and
appropriate clinical care at the end of life.
Axelsson, B & Christensen, SB 1996. Place of death correlated to sociodemographic
factors. A study of 203 patients dying of cancer in a rural Swedish county in 1990.
Palliative Medicine; 10(4): 329-335.
Badrakalimuthu, V., & Barclay, S. (2014). Do people with dementia die at their
preferred location of death? A systematic literature review and narrative synthesis.
Age and Ageing, 43,13-19.
Bakitas, M., Ahles, T.A., Skalla, K., Brokaw, F.C., Byock, I., Hanscom, B., et al.
(2008). Proxy Perspectives about End-of-Life Care for Person’s with Cancer.
Cancer; 112(8): 1854–1861.
Bakitas, M. A., Elk, R., Astin, M., Ceronsky, L., Clifford, K. N., Dionne-Odom, J. N.,
. . . Smith, T. (2015). Systematic Review of Palliative Care in the Rural Setting.
Cancer Control, 22(4), 450-464.
Bausewein, C. (2017). Beyond cancer: challenges and opportunities for early palliative
care engagement. Plenary session, 15th World Congress of the European Association
of Palliative Care; Madrid, Spain.
235
Becker, H., Gans, H., Newman, K., & Vaughan, D. (2004). On the value of
Ethnography: sociology and public policy, a dialogue. The Annals of the American
Academy of Political and Social Science, 595, 264-276.
Beckstrand, R.L., Giles, V.C., Luthy, K.E., Callister, L.C., & Heaston, S. (2012).The
last frontier: rural emergency nurses' perceptions of end-of-life care obstacles.
Journal of Emergency Nursing, 38(5), E15-25.
Beckstrand, R.L., Rohwer, J., Luthy, K.E., Macintosh, J.L., & Rasmussen, R.J. (2017).
Rural Emergency Nurses' End-of-Life Care Obstacle Experiences: Stories from the
Last Frontier. Journal of Emergency Nursing, 43(1), 40-48.
Biggs, D. D. (2015). "Comfort, comfort my people": A qualitative study examining
dying persons' and their caregivers' experience of spiritual care as an aid to having a
good death in grey and bruce counties, Ontario. Dissertation Abstracts International
Section A: Humanities and Social Sciences, 75(11-A(E)), No Pagination Specified.
Black, H., Waugh, C., Munoz-Arroyo, R., Carnon, A., Allan, A., Clark, D., … Isles, C.
(2016). Predictors of place of death in South West Scotland 2000–2010:
Retrospective cohort study. Palliative Medicine, 30(8), 764–771.
http://doi.org/10.1177/0269216315627122
Brameld, K., Spilsbury, K., Rosenwax, L., Murray, K., & Semmens, J. (2017). Issues
using linkage of hospital records and death certificate data to determine the size of a
potential palliative care population. Palliative Medicine, 31(6), 537–543.
http://doi.org/10.1177/0269216316673550
Brazil, K., Kaasalainen, S., Williams, A., & Rodriguez, C. (2013). Comparing the
experiences of rural and urban family caregivers of the terminally ill. Rural and
Remote Health, 13(1).
Brazil, K., Kaasalainen, S., Williams, A., & Dumont, S. (2014). A comparison of
support needs between rural and urban family caregivers providing palliative care.
American Journal of Hospice and Palliative Care, 31(1), 13-19.
Broad, J.B., Gott, M., Kim, H., Boyd, M., Chen, H., & Connolly, M.J. (2012). Where do
people die? An international comparison of the percentage of deaths occurring in
hospital and residential aged care settings in 45 populations, using published and
available statistics. International Journal of Public Health, 58, 257-267.
Bullen, T., Maher, K, Rosenberg, J., & Smith, B. (2014). Establishing research in a
palliative care clinical setting: perceived barriers and implemented strategies.
Applied Nursing Research, 27(1), 78-83.
Buikstra, E., Pearce, S., Hegney, D., & Fallon, T. (2006). SEAM - improving the
quality of palliative care in regional Toowoomba, Australia: lessons learned. Rural
and Remote Health, 6(1), 415.
Burge, F. (2005). Where a cancer patient dies: effect of rural residency. Journal of
Rural Health, 21(3), 233-238.
236
Burge, F., Lawson, B., & Johnston, G. (2003). Trends in the place of death of cancer
patients, 1992-1997. Canadian Medical Association Journal, 168(3), 265-270.
Burns, C.M., Dal Grande, E., Tieman, J., Abernethy, A.P., & Currow, D.C. (2015).
Who provides care for people dying of cancer? A comparison of a rural and
metropolitan cohort of South Australian bereaved population study. Australian
Journal of Rural Health, 23(1), 24-31.
Byock, I., Curtis, J., & Patrick, D. (2001). Improving end-of-life experience and care in
the community: a conceptual framework. Journal of Pain and Symptom
Management, 22(3), 759-772.
Cairns, W. (2007). A Short history of palliative medicine in Australia [online]. Cancer
Forum, 31(1).
http://search.informit.com.au/documentSummary;dn=443061458381634;res=IE
LHEA> ISSN: 0311-306X>. [Viewed 21 January 2017].
Carroll, D. (1998). An audit of place of death of cancer patients in a semi-rural Scottish
practice. Palliative Medicine, 12(1), 51-53.
Cartwright, J., & Kayser-Jones, J. (2003).End-of-life care in assisted living facilities:
perceptions of residents, families, and staff. Journal of Hospice & Palliative Nursing,
5(3), pp. 143-151.
Casarett, D. (2005). Ethical considerations in end-of- life care and research. Journal of
Palliative Medicine, 8 (Supp 1), S148-160.
Castillo-Guzman, S., Palacios-Rios, D., Nava-Obregon, T.A., Torres-Perez, J.F. &
Gonzalez-Santiago, O. (2013). Home versus hospital mortality from cancer in
Mexico (1999-2009). American Journal of Hospice and Palliative Care, 30(3), 249-
252.
Castleden, H., Crooks, V., Hanlon, N., & Schuurman, N. (2010). “It’s not necessarily
the distance on the map…”: Using place as an analytic tool to elucidate geographical
issues central to rural palliative care. Health and Place, 16(2), 284-290.
Catalan-Fernandez, J.G., Pons-Sureda, O., Recober-Martinez, A., Avella-Mestre, A.,
Carbonero-Malberti, J. M., Benito-Oliver, E., Garau-Llinas, I. (1991). Dying of
cancer. The place of death and family circumstances. Medical Care, 29(9), 841-852.
Champion, N., Medigovich, K., & Morrison, P. (2015). Exploring why some terminally
ill people die in hospital when home was their preferred choice. Australian Journal
of Advanced Nursing, 33(2), 6-12.
Chi, N.C., & Demiris, G. (2015). A systematic review of telehealth tools and
interventions to support family caregivers. Journal of Telemedicine and Telecare,
21(1), 37–44.
Choi, J.Y., Chae, Y.M., Lee, C.G., Kim, S.Y., Lee, S.W., et al. (2005). Factors
influencing preferences for place of terminal care and of death among cancer patients
and their families in Korea. Supportive Care in Cancer, 13(8), 565-572.
237
Ciemins, E.L., Brant, J., Kersten, D., Mullette, E., & Dickerson, D. (2015). A quality
analysis of patients and family perspectives of palliative care. Journal of Palliative
Medicine, 18(3), 282-285.
Cohen, J., Bilsen, J., Addington-Hall, J., et al. (2008). Population based study of dying
in hospital in six European countries. Palliative Medicine, 22, 702-710.
Cohen, J., Houtter, D., Onwuteaka-Philipsen, B., & Miccinesi, G., Addington-Hall, J.,
Kaasa, S., et al. (2010). Which patients with cancer die at home? A study of six
European countries using death certificate data. Journal of Clinical Oncology,
28(13), 2267-73.
Cohen, J., Pivodic, L., Miccinesi, G., et al. (2015). International study of the place of
death of people with cancer: a population-level comparison of 14 countries across 4
continents using death certificate data. British Journal of Cancer, 113, 1397–1404.
Collier, C., Phillips, J.L., & Iedema, R. (2015). The meaning of home at the end of life:
a video-reflexive ethnography study. Palliative Medicine, 29(8), 695-702.
Colquhoun, H.L., Levac, D., O’Brien, K.K., et al. (2014). Scoping reviews: time for
clarity in definition, methods and reporting. Journal of Clinical Epidemiology,
67(12), 1291-1294.
Costantini, M., Balzi, D., Garronec, E., Orlandini, C., Parodi, S., Vercelli, M., & Bruzzi,
P. (2000). Geographical variations of place of death among Italian communities
suggest an inappropriate hospital use in the terminal phase of cancer disease. Public
Health, 114(1), 15-20.
Cottrell, L., & Duggleby, W. (2016). The “good death”: an integrative literature review.
Palliative and Supportive Care, 14(6): 686-712.
Counts, D.A., & Counts, D. (2004). The good, the bad, and the unresolved death in
Kaliai. Social Science and Medicine, 58(5), 887-897.
Crotty, M. (1998). The foundations of social research: Meaning and perspective in the
research process. London, Thousand Oaks, California, Sage Publications.
Cruickshank, S., Adamson, E., Logan, J., & Brackenridge, K. (2010). Using syringe
drivers in palliative care within a rural community setting: capturing the whole
experience. International Journal of Palliative Nursing, 16(3), 126-132.
Crawford, G. (2000). Palliative care on Kangaroo Island. Australian Journal of Rural
Health, 8, 35-40.
Curtis, J.R., Patrick, D., Engelberg, R., Norris, K., Asp, C., & Byock, I. (2002). A
measure of the quality of dying and death: initial validation using after-death
interviews with family members. Journal of Pain and Symptom Management, 24(1),
17-30.
Dale, M.J. & Johnston, B. (2011). An exploration of the concerns of patients with
inoperable lung cancer. International Journal of Palliative Nursing, 17(6), 285-290.
238
Darer, J.D., Clarke, D.N., Sees, A.C., et al. (2015). Assessing the quality of death and
dying in an integrated health care system in rural Pennsylvania. Journal of Pain and
Symptom Management, 50(3), 343-349.e6.
Daudt, H. M., van Mossel, C., & Scott, S. J. (2013). Enhancing the scoping study
methodology: a large, inter-professional team’s experience with Arksey and
O’Malley’s framework. BMC Medical Research Methodology, 13, 48.
http://doi.org/10.1186/1471-2288-13-48
Davies, A., Todd, J., Bailey, F., Gregory, A., & Waghorn, M. (2016). Good
concordance between patients and their non-professional carers about factors
associated with a ‘good death’ and other important end-of-life decisions. BMJ
Supportive and Palliative Care, [Epub ahead of print].
De Raeve, L., (1994). Ethical issues in palliative care research. Palliative Medicine,
8(4), 298-305.
Dekker, A.M., Amon, J.J., Le Roux, K.W., & Gaunt, C.B. (2012). “What is killing me
most”: Chronic pain and the need for palliative care in the Eastern Cape, South
Africa. Journal of Pain and Palliative Care Pharmacology, 26(4), 334-340.
Dembinsky, M. (2014). Exploring Yamatji perceptions and use of palliative care: an
ethnographic study. International Journal of Palliative Nursing, 20(8), 387-393.
Dept. of Health (UK). (2008). ‘End of life care strategy: promoting high quality care for
all adults at the end of life’. LONDON, HMSO, www.dh.gov.uk/publications.
Despres, C. (1991). The meaning of home: literature review and directions for future
research and theoretical development. Journal of Architectural and Planning
Research, 8(2), 96-115.
Devik, S.A., Enmarker, I., Wiik, G.B., & Hellzen, O. (2013). Meanings of being old,
living on one's own and suffering from incurable cancer in rural Norway. European
Journal of Oncology Nursing, 17(6), 781-7.
Devik, S.A., Hellzen, O., & Enmarker, I. (2015). "Picking up the pieces" - Meanings of
receiving home nursing care when being old and living with advanced cancer in a
rural area. International Journal of Qualitative Studies on Health and Well-being,
10:28382. doi:10.3402/qhw.v10.28382 [online].
Devik, S.A., Hellzen, O., & Enmarker, I. (2016). Bereaved family members’
perspectives on suffering among older rural cancer patients in palliative home
nursing care: A qualitative study. European Journal of Cancer Care, 1-9.
doi:10.1111/ecc.12609 [online].
Diaz, A., Whop, L.J., Valery, P.C., et al. (2015). Cancer outcomes for Aboriginal and
Torres Strait Islander Australians in rural and remote areas. Australian Journal of
Rural Health, 23(1), 4-18.
239
Dilger, H. (2008). “We are all going to die”: kinship, belonging, and the morality of
HIV/AIDS-related illnesses and deaths in rural Tanzania. Anthropological Quarterly,
81(1), 207-232.
Downing, J., & Jack, B. (2012). End-of-life care in rural areas: What’s the difference?
Current Opinion in Supportive and Palliative Care, 6(3), 391-397.
Doyle, D., Hanks, G.W., & MacDonald, N. (eds.) (1999). Oxford textbook of palliative
medicine. 2nd Ed. University Press, Oxford. Foreword by Cicely Saunders, pp. v-ix.
Duggleby, W.D., Penz, K.L., Goodridge, D.M., Wilson, D.M., et al. (2010). The
transition experience of rural older persons with advanced cancer and their families:
a grounded theory study. BMC Palliative Care, 9,5 [online].
Duggleby, W.D., Penz, K., Leipert, B.D., Wilson, D.M., Goodridge, D., & Williams, A.
(2011). 'I am part of the community but...' The changing context of rural living for
persons with advanced cancer and their families. Rural and Remote Health, 11(3),
1733 [online].
Duggleby, W.D., Williams, A., & Holstlander, L. (2014). Hope of rural women
caregivers of persons with advanced cancer: guilt, self-efficacy and mental health.
Rural and Remote Health, 14, 2561 [online].
Easom, L., Galatas, S., & Warda, M. (2006). End-of-life care: an educational
intervention for rural nurses in southeastern USA. International Journal of Palliative
Nursing, 12(11), 526-534.
Earle, C.C., Park, E.C., Lai, B., Weeks, J., Ayanian, J.Z., & Block, S. (2003).
Identifying potential indicators of the quality of end-of-life cancer care from
administrative data. Journal of Clinical Oncology, 21, 1133-1138.
Equator Network 2009. Preferred Reporting Items for Systematic Reviews and Meta-
Analysis extension for Scoping Reviews (PRISMA-ScR); viewed 27 February 2016.
http://www.equator-network.org/library/reporting-guidelines-under-
development/#55; http://www.equator-network.org/wp-
content/uploads/2009/02/Executive-summary_ScR_Dec-9.pdf.
Escobar Pinzon, L.C., Claus, M., Zepf, K.I., Letzel, S., Fischbeck, S., & Weber, M.
(2011a). Preference for place of death in Germany. Journal of Palliative Medicine,
14(10), 1097-1103.
Escobar Pinzon, L.C., Weber, M., Claus, M., Fischbeck, S., et al. (2011b). Factors
influencing place of death in Germany. Journal of Pain and Symptom Management,
41(5), 893-903.
Ewing, G., Rogers, M., Barclay, S., McCabe, J., Martin, A., & Todd, C. (2004).
Recruiting patients into a primary care based study of palliative care: why is it so
difficult? Palliative Medicine, 18(5), 452-459.
240
Felt, D.H., Early, J.L., & Welk, T.A. (2000). Attitudes, values, beliefs, and practices
surrounding end-of-life care in selected Kansas communities. American Journal of
Hospital and Palliative Care, 17(6), 401-6.
Fink, R.M., Oman, K.S., Youngwerth, J., & Bryant, L.L. (2013). A palliative care needs
assessment of rural hospitals. Journal of Palliative Medicine, 16(6), 638-644.
Fischer, D., Burgener, S., Kavanaugh, K., Ryan, C., & Keenan, G. (2011). Conducting
research with end-of-life populations: overcoming recruitment challenges when
working with clinical agencies. Applied Nursing Research, 25(4), 258-263.
Foreman, L.M., Hunt, R.W., Luke, C.G., & Roder, D.M. (2006). Factors predictive of
preferred place of death in the general population of South Australia. Palliative
Medicine, 20(4), 447-53.
Fukui, S., Yoshiuchi, K., Fujita, J., Sawai, M., & Watanabe, M. (2011). Japanese
people’s preference for place of end-of-life care and death: a population-based
nationwide survey. Journal of Pain and Symptom Management, 42(6), 882-92.
Gatrell, A.C., Harman, J.C., Francis, B.J., Thomas, C., Morris, S.M., & McIllmurray,
M. (2003). Place of death: an analysis of cancer deaths in part of North West
England. Journal of Public Health Medicine, 25(1), 53-58.
Gessert, C., Waring, S., Bailey–Davis, L., Conway, P., Roberts, M., & Van Womer, J.
(2015). Rural definition of health: a systematic literature review. BMC Public
Health; 15:378 [online].
Gibbons, J., Reid, C., Bloor, S., Burcomb, M., McCoubrie, R., & Forbes, K. (2013).
Overcoming barriers to recruitment in care of the dying research in hospitals. Journal
of Pain and Symptom Management, 45(5), 859-867.
Gold Standard Framework, 4th ed. 2011.
http://www.goldstandardsframework.org.uk/cd-
content/uploads/files/General%20Files/Prognostic%20Indicator%20Guidance%20Oc
tober%202011.pdf
Gomes, B., & Higginson, I.J. (2006). Factors influencing death at home in terminally ill
patients with cancer: systematic review. British Medical Journal, 332(7540), 515-21.
Gomes, B., & Higginson, I. (2008). Where people die (1974-2030): past trends, future
projections and implications for care. Palliative Medicine, 22, 33-41.
Gomes, B., Calanzani, N., Gysels, M., Hall, S., & Higginson, I. (2013). Heterogeneity
and changes in preferences for dying at home: a systematic review. BMC Palliative
Care, 12:7. doi:10.1186/1472-684X-12-7. [Online].
Gomez-Batiste, X., Martinez-Munoz, M., Blay, C., Amblas, J., Villa, L., et al. (2014).
Prevalence and characteristics of patients with advances chronic conditions in need
of palliative care in the general population: A cross-sectional study. Palliative
Medicine, 28(4), 302-311.
241
Goodridge, D., & Duggleby, W. (2010a). Using a quality framework to assess rural
palliative care. Journal of Palliative Care, 26(3), 141-150.
Goodridge, D., Lawson, J., Rennie, D., & Marciniuk, D. (2010b). Rural/urban
differences in health care utilization and place of death for persons with respiratory
illness in the last year of life. Rural and Remote Health, 10(2), 1349. [Online].
Gott, M., Allen, R., Moeke-Maxwell, T., Gardiner, C., & Robinson, J. (2015). ‘No
matter what the cost’: A qualitative study of the financial costs faced by family and
whanau caregivers within a palliative care context. Palliative Medicine, 29(6), 518-
528.
Gott, M., Frey, R., Robinson, J., Boyd, M., O’Callaghan, A., Richards, N., & Snow, B.
(2013). The nature of, and reasons for, ‘inappropriate’ hospitalisations among
patients with palliative care needs: A qualitative exploration of the views of
generalist palliative care providers. Palliative Medicine, 27(8), 747–756.
Gott, M., Seymour, J., Bellamy, G., Clark, D., & Ahmedzai, S. (2004). Older people’s
views about home as a place of care at the end of life. Palliative Medicine, 18(5),
460-467.
Grant, E., Murray, S.A., Grant, A., & Brown, J. (2003). A good death in rural Kenya? -
Listening to Meru patients and their families talk about care needs at the end of life.
Journal of Palliative Care, 19(3), 159-167.
Grant, L., Brown, J., Leng, M., Bettega, N., & Murray, S.A. (2011). Palliative care
making a difference in rural Uganda, Kenya and Malawi: Three rapid evaluation
field studies. BMC Palliative Care, 10(8), 1-10.
Grbich, C. (1999). Qualitative research in Health: an introduction. Reprint 2003,
London; Thousand Oaks, Calif.: Sage Publications.
Grbich, C. (2015). Course notes for Qualitative methodologies, data collection and
analytic approaches. ACSPRI. ANU, Canberra 19-23 January 2015. p. 40.
Gu, X., Cheng, W., Cheng, M., Liu, M., & Zang, Z. (2015). The preference of place of
death and its predictors among terminally ill patients with cancer and their caregivers
in China. American Journal of Hospice and Palliative Medicine, 32(8), 835-840.
Guillemin, M., & Gillam, L. (2004). Ethics, reflexivity, and ethically important
moments in research. Qualitative Inquiry, 10(2), 274-5.
Gysels, M., Evans, C., & Higginson, I. (2012). Patient, caregiver, health professionals
and researcher views and experiences of participating in research at the end of life: a
critical interpretive synthesis of the literature. BMC Medical Research Methodology,
12:123 [open access].
Hales, S., Zimmerman, C., & Rodin, G. (2010). The quality of dying and death: a
systematic review of measures. Palliative Medicine, 24(2), 127-144.
242
Hansen, L., Cartwright, J.C., & Craig, C.E. (2012). End-of-life care for rural-dwelling
older adults and their primary family caregivers. Research in Gerontological
Nursing, 5(1), 6-15.
Hart, L.G., Larson, E.H., & Lishner, D.M. (2005). Rural definitions for health policy
and research. Critical concepts for reaching populations at risk. American Journal of
Public Health, 95(7), 1149-1155.
Hatcher, I., Harms, L., Walker, B., et al. (2014). Rural palliative care transitions from
home to hospital: Carers' experiences. Australian Journal of Rural Health, 22(4),
160-164.
Hawker, S., Payne, S., Kerr, C., Hardey, M., & Powell, J. (2002). Appraising the
evidence: reviewing disparate data systematically. Qualitative Health Research, 12,
1284–1299.
Herce, M., Elmore, S.N., Kalanga, N., et al. (2014). Assessing and responding to
palliative care needs in rural Sub-Saharan Africa: results from a model intervention
and situation analysis in Malawi. Plos One, 9(10):17 [online].
Herd, E. (1990). Terminal Care in a semi- rural area. British Journal of General
Practice, 40, 248-251.
Higginson, I.J., Reilly, C.C., Bajwah, S., Maddocks, M., et al. 2017. Which patients
with advanced respiratory disease die in hospital? A 14-year population-based study
of trends and associated factors. BMC Medicine, 15(19):1-12 [online].
Hoare, S., Morris, Z.S., Kelly, M.P., Kuhn, I., &Barclay, S. (2015). Do patients want to
die at home? A systematic review of the UK literature, focused on missing
preferences for place of death. PLoS One, 10(11) [online].
Holdsworth, L.M. (2015). Bereaved carers’ accounts of the end of life and the role of
care providers in a ‘good death’: A qualitative study. Palliative Medicine, 29(9), 834-
841.
Horsfall, D., Leonard, R., Noonan, K., & Rosenberg, J. (2013). Working together-apart:
exploring the relationships between formal and informal care networks for people
dying at home. Progress in Palliative Care, 21(6), 331-336.
Houttekier, D., Cohen, J., Surkyn, J., et al. (2011). Study of recent and future trends in
place of death in Belgium using death certificate data. BMC Public Health, 11,
228[online].
Howarth, G. (2007). Death and dying: a sociological introduction. Cambridge, England:
Polity Press.
Howat, A., Veitch, C., & Cairns, W. (2007). A retrospective review of place of death of
palliative care patients in regional north Queensland. Palliative Medicine, 21(1), 41-
47.
243
Howell, D., Marshall, D., Brazil, K., et al. (2011). A shared care model pilot for
palliative home care in a rural area: impact on symptoms, distress, and place of
death. Journal of Pain and Symptom Management, 42(1), 60-75.
Hughes, P.M., Ingleton, M.C., & Clark, D. (2004). Providing cancer and palliative care
in rural areas: a review of patient and carer needs. Journal of Palliative Care, 20(1),
44-49.
Hunt, R.W., Fazekas, B.S., Luke, C.G., & Roder, D.M. (2001). Where patients with
cancer die in South Australia, 1990-1999: A population-based review. Medical
Journal of Australia, 175(10), 526-529.
Huy, T.Q., Johansson, A., & Long, N.H. (2007). Reasons for not reporting deaths: a
qualitative study in rural Vietnam. World Health and Population, 9(1), 14-23.
Jayaraman, J., & Joseph, K. S. (2013). Determinants of place of death: a population-
based retrospective cohort study. BMC Palliative Care, 12, 19.
doi:http://dx.doi.org/10.1186/1472-684X-12-19
Joanna Briggs Institute. (2015). Reviewers’ manual: methodology for JBI scoping
reviews. Viewed 12th February 2016. www.joannabriggs.org.
Joarder, T., Cooper, A., & Zaman, S. (2014). Meaning of death: an exploration of
perception of elderly in a Bangladeshi village; Journal of Cross Cultural
Gerontology, 29(3), 299-314.
Johnston, B.M., Milligan, S., Foster, C., & Kearney, N. (2012). Self-care and end of life
care, patients' and carers' experience: a qualitative study utilising serial triangulated
interviews. Supportive Care in Cancer, 20(8), 1619-27.
Johnston, N., Lovell, C., Liu, W-M., & Chapman M. (2016). Normalising and planning
for death in residential care: findings from a qualitative focus group study of a
specialist palliative care intervention. Supportive and Palliative Care; published
online. doi:10.1136/bmjspcare-2016-001127.
Kaasalainen, S., Brazil, K., Williams, A., et al. (2014). Nurses’ experiences providing
palliative care to individual living in rural communities: aspects of the physical
residential setting. Rural and Remote Heath, 14:2728 [online].
Kearl, M. (1996). Dying well: The unspoken dimension of aging well. American
Behavioural Scientist, 39(3), 336–360.
Kellehear, A. (2008). A social history of dying. Cambridge University Press, New
York.
Kelly, L., Linkewich, B., Cromarty, H., Pierre-Hansen, N.S., Antone, I., & Gilles, C.
(2009). Palliative care of first nations people: a qualitative study of bereaved family
members. Canadian Family Physician, 55(4), 394-395.
244
Kirby, S., Barlow, V., Saurman, E., Lyle, D., Passey, M., & Currow, D. (2016). Are
rural and remote patients, families and caregivers needs in life-limiting illness
different from those of urban dwellers? A narrative synthesis of the evidence.
Australian Journal of Rural Health, 24, 289-299.
Kleinman, A. (1988). The illness narratives: suffering, healing and the human condition.
United States: Basic Books.
Knight, D.A. (2014). What constitutes a good death? A clergy perspective. Rural
Theology, 12(1), 2-16.
Kuper, A., Lingard, L., & Levinson, W. (2008). Critically appraising qualitative
research. British Medical Journal, 337:a1035.
Lavergne, M.R., Lethbridge, L., Johnston, G., Henderson, D., D’Intino, A.F., &
McIntyre, P. (2015). Examining palliative care program use and place of death in
rural and urban contexts: a Canadian population-based study using linked data. Rural
and Remote Health; 3134 [online].
Lazenby, M., Ma, T., Moffat, H.J., Funk, M., Knobf, M.T., & McCorkle, R. (2010).
Influences on place of death in Botswana. Palliative and Supportive Care, 8(2), 177-
85.
Lazenby, J., & Olshevski, J. (2012). Place of death in Botswana’s oldest old. Omega:
Journal of Death and Dying, 65(3), 173-187.
Levac, D., Colquhoun, H., & O’Brien, K.K. (2010). Scoping studies: advancing the
methodology. Implementation Science, 5, 69. doi: 10.1186/1748-5908-5-69.
Lin, H.C., &Lin, C.C. (2007). A population based study on the specific locations of
cancer deaths in Taiwan, 1997-2003. Supportive Care in Cancer, 15(12), 1333-1339.
Lockie, S.J., Bottorff, J.L., Robinson, C.A., &Pesut, B. (2010). Experiences of rural
family caregivers who assist with commuting for palliative care. Canadian Journal of
Nursing Research, 42(1), 74-91.
Lopez-Campos, J.L., Ruiz-Ramos, M., Mendez, C., & Garcia-Leon, J. (2013).
Characteristics of subjects who dies of chronic obstructive pulmonary disease in
Andalusia in 2010 and 2011. Journal of Palliative Medicine, 16(12), 1610-13.
Lynch, S. (2013). Hospice and palliative care access issues in rural areas. American
Journal of Hospice and Palliative Medicine, 30(2), 172-177.
MacLeod, R., Johnson, C., & Yule, S. (2015). Making it easier for people to die at
home- an innovative programme in New South Wales, Australia. European Journal
of Palliative Care, 22(1), 26-29.
May, P. (2017). Measuring palliative care’s effect on healthcare costs. Oral
presentation, parallel session, 15th World Congress of the European Association of
Palliative Care, Madrid, Spain.
245
McCall, K., & Rice, A.M. (2005). What influences decisions around the place of care
for terminally ill cancer patients? International Journal of Palliative Nursing, 11(10),
541-547.
McGrath, P. (2006). Exploring Aboriginal peoples’ experiences of relocation for
treatment during end-of-life care. International Journal of Palliative Nursing, 12(3),
102- 108.
McGrath, P. (2007). 'I don't want to be in that big city; this is my country here':
Research findings on Aboriginal peoples' preference to die at home. Australian
Journal of Rural Health, 15(4), 264-268.
McNamara, B. (2004). Good enough death: autonomy and choice in Australian
palliative care. Social Science and Medicine, 58, 929–938.
McNamara, B., Rosenwax, L., & Holman, C.D.J. (2006). A method for defining and
estimating the palliative care population. Journal of Pain and Symptom Management,
32(1), 5-12.
McNamara, B., & Rosenwax, L. (2007). Factors affecting place of death in Western
Australia. Health and Place, 13, 356-367.
Meier, E., Gallegos, J., Montross-Thomas, L., Depp, C., Irwin, S., & Jeste, D. (2016).
Defining a Good Death (Successful Dying): Literature Review and a Call for
Research and Public Dialogue. American Journal of Geriatric Psychiatry, 24(4), 261–
271.
Menec, V., Nowicki, S., & Kalischuk, A. (2010). Transfers to acute care hospital at the
end of life: do rural/remote regions differ from urban regions? Rural and Remote
Health, 10, 1281 [Online].
Mitchell, GK. (2002). How well do general practitioners deliver palliative care? A
systematic review. Palliative Medicine, 16(6), 457-64.
Mitchell, GK., Johnson, C.E., Thomas, K., &Murray, S.A. (2010). Palliative care
beyond that for cancer in Australia. Medical Journal of Australia, 193(2), 124-126.
Mixer, S., Fornehed, M.L., Varney, J., & Lindley, L.C. (2014). Culturally congruent
end-of-life care for rural Appalachian people and their families. Journal of Hospice
and Palliative Nursing, 16(8), 526-535.
Miyajima, K., Fujisawa, D., Yoshimura, K., Ito, M., et al. (2014). Association between
quality end-of-life care and possible complicated grief amongst bereaved family
members. Journal of Palliative Medicine, 19(9), 1025-1031.
Moher, D., Liberati, A., Tetzlaff, J., et al. (2009). Preferred reporting items for
systematic reviews and meta-analyses: the PRISMA statement. Annals of Internal
Medicine, 151(4), 264-270.
Monaro Regional Health Services (MRHS). (2015). Health care services plan 2015-
2019. Southern NSW Local Health District.
246
Morris, S., & Thomas, C. (2005). ‘Placing the dying body: emotional, situational and
embodied factors in preferences for place of final care and death in cancer’, in
Davison J, Bondi L & Smith M (eds). Emotional Geographies. Ashgate; Chapter 2.
Murray, S.A., Kendall, M., Boyd K., & Sheikh, A. (2005). Illness trajectories and
palliative care. British Medical Journal, 330(7498), 1007–1011.
doi:10.1136/bmj.330.7498.1007
Mystakidou, K., Parpa, E., Tsilika, E., et al. (2009). Where do cancer patients die in
Greece? A population-based study on the place of death in 1993 and 2003. Journal of
Pain and Symptom Management, 38(2), 309-314.
New South Wales Agency for Clinical Innovation (NSW ACI) 2015. ‘Palliative and
end-of-life care. A blueprint for improvement’; viewed 20 March 2017.
<https://www.aci.health.nsw.gov.au/palliative-care-blueprint>
NSW Government 2012. Aging Strategy, Department of Family and Community
Services, Office for Ageing; viewed 21 January 2017.
<https://www.facs.nsw.gov.au/__data/assets/pdf_file/0005/382946/1282_ADHC_NS
W-AgeingStrategy_WEB.pdf >.
NSW Ministry of Health 2012. The NSW Government plan to increase access to
palliative care 2012-2016; viewed 28 November 2015.
<http://www.health.nsw.gov.au/palliativecare/Publications/Palliative-Care-Plan-
2012-2016.pdf >.
NSW Ministry of Health 2013. Last days of life home support services; viewed 20
March 2017. <http://www.health.nsw.gov.au/palliativecare/Pages/last-days-of-life-
home-support-services.aspx>
NSW Government Media Release 2017. Budget 2017-2018; viewed 26 June 2017.
< https://www.nsw.gov.au/your-government/the-premier/media-releases-from-the-
premier/nsw-budget-record-100m-palliative-care-funding/>.
O’Connor, M., & Lee-Steere, R. (2006). General Practitioners’ attitudes to palliative
care: A Western Australian rural perspective. Journal of Palliative Medicine, 9(6),
1271-1281.
Ostertag, S.G., &Forman, W.B. (2008). End-of-life care in Hancock County, Maine: a
community snapshot. American Journal of Hospice and Palliative Medicine, 25(2),
132-138.
Palliative Care Australia 2003. Palliative Care Service Provision in Australia: A
Planning Guide. www.pallcare.org.au.
Palliative Care Council South Australia 2016; viewed 19 September 2016.
http://www.pallcare.asn.au/PalliativeCareSA/History-of-palliative-care
Papke, J., & Koch, R. (2007). Places of death from cancer in a rural location.
Onkologie, 30(3), 105-108.
247
Patrick, D., Engelberg, R., &Curtis, J.R. (2001). Evaluating the quality of dying and
death. Journal of Pain and Symptom Management, 22(3), 717-726.
Patton, MQ. (2002). Qualitative research and evaluation methods, 3rd ed. Thousand
Oaks, California: Sage Publication.
Payne, S., Kerr, C., Hawker, S., Seamark, D., Davis, C., Roberts, H., et al. (2004).
Community hospitals: an under-recognized resource for palliative care. Journal of the
Royal Society of Medicine, 97, 428–431.
Payne, S., Hawker, S., & Kerr, C. (2007). Experiences of end-of-life care in community
hospitals. Health and Social Care in the Community, 15(5), 494-501.
Pereira, G. (2005). Palliative care in the hinterlands: A description of existing services
and doctors’ attitudes. Australian Journal of Rural Health, 13, 343–347.
Pesut, B., Robinson, C.A., Bottorff, J.L., Fyles, G., & Broughton, S. (2010). On the road
again: Patient perspectives on commuting for palliative care. Palliative and
Supportive Care, 8(2), 187-195.
Pesut, B., Bottorff, J.L., & Robinson, C.A. (2011). Be known, be available, be mutual:
A qualitative ethical analysis of social values in rural palliative care. BMC Medical
Ethics, 12:19 [online].
Pesut, B., Robinson, C.A., & Bottorff, J.L. (2014). Among neighbors: an ethnographic
account of responsibilities in rural palliative care. Palliative and Supportive Care,
12(2), 127-138.
Peters, M.D.J., Godfrey, C.M., Khalil, H., McInerney, P., Parker, D., & Soares, C.B.
(2015). Guidance for conducting systematic scoping reviews. International Journal of
Evidence Based Healthcare, 13, 141-146.
Phillips, J.L., & Currow, D.C. (2017). Would reframing aged care facilities as a
‘hospice’ instead of ‘home’ enable older people to get the care they need? Collegian,
24, 1-2.
Plowman, S. (2007). Cooma-Monaro Shire – Thematic History 1823-1945; viewed 16
January 2017. http://pandora.nla.gov.au/pan/86625/20080704-1100/CMSHistory.pdf
Plowman, S., Dixon, T., & Rushton, B. (eds.) (1999). Cooma-Monaro 150 years on
committee, ed. Cooma-Monaro Historical Society. South-East Printing, Cooma.
Productivity Commission 2013. An Ageing Australia: Preparing for the Future,
Commission Research Paper Overview, Canberra; viewed 21 January 2017.
<http://www.pc.gov.au/research/completed/ageing-australia>.
Proulx, K., & Jacelon, C. (2004). Dying with dignity: The good patient versus the good
death. American Journal of Hospice and Palliative Care, 21(2), 116-120.
248
Rainsford, S., Bullen, T., Rosenberg, J.P. (2015). Challenges of recruiting hospice in-
patients with advanced cancer to research: reflections on a delirium study. Progress
in Palliative Care, 23(1), 24-28.
Rainsford, S., MacLeod, R.D., Glasgow, N.J., Wilson, D.M., Phillips, C.B, & Wiles,
R.B. (2016). Rural residents’ perspectives on the rural ‘good death’: A scoping
review. Health and Social Care in the Community. Epub ahead of print.
Rainsford, S., MacLeod, R.D. & Glasgow, N.J. (2016). Place of death in rural palliative
care: A systematic review. Palliative Medicine, 30(8), 745–763.
Rainsford, S., MacLeod, R.D., Glasgow, N.J., Phillips, C.B., Wiles, R.B, & Wilson,
D.M. (2017). Rural end-of-life care from the experiences and perspectives of patients
and family caregivers: A systematic literature review. Palliative Medicine, 31(10);
895-912.
Raisio, H., Vartiainen, P., & Jekunen, A. (2015). Defining a Good Death: A deliberative
democratic view. Journal of Palliative Care, 31(3), 158 –165.
Reese, D.J., Smith, M.R., Butler, C., Shrestha, S., & Erwin, D.O. (2014). African
American client satisfaction with hospice: a comparison of primary caregiver
experiences within and outside of hospice. American Journal of Hospice and
Palliative Care, 31(5), 495-502.
Revier, S.S., Meiers, S.J., & Herth, K.A. (2012). The lived experience of hope in family
caregivers caring for a terminally ill loved one. Journal of Hospice and Palliative
Nursing, 14(6), 438-446.
Rice, P.L. & Ezzy, D. (1999). Rigour, Ethics and Sampling. Qualitative Research
Methods: A health focus. Melbourne: Oxford University Press. P41-50.
Robinson, C., Pesut, B., Botorff, J., Mowry, A., Broughton, S., & Fyles, G. (2009).
Rural Palliative Care: A Comprehensive Review. Journal of Palliative Medicine,
12(3), 253-258.
Robinson, C., Pesut, B., & Bottorff, J. (2010). Issues in rural palliative care: views from
the countryside. Journal of Rural Health, 26(1), 78-84.
Robinson, J., Gott, M., Gardiner, C., &Ingleton, C. (2012). A qualitative study
exploring the benefits of hospital admissions from the perspectives of patients with
palliative care needs. Palliative Medicine, 8,703-710.
Robinson, J., Gott, M., Gardiner, C., & Ingleton C. (May, 2017). Hospital admissions in
palliative care: benefit and burden. Poster 15th World Congress of the European
Association of Palliative Care, Madrid, Spain.
Ross, C., & Cornbleet, M. (2003). Attitudes of patients and staff to research in a
specialist palliative care unit. Palliative Medicine, 17, 491-497.
249
Rowland, C., Hanratty, B., Pilling, M., van den Berg, B., & Grande, G. (2017). The
contributions of family care-givers at end of life: A national post-bereavement census
survey of cancer carers’ hours of care and expenditures. Palliative Medicine, 31(4),
346-355.
Sandsdalen, T., Hov, R., Hoye, S., Rystedt, I., & Wilde-Larsson, B. (2015). Patients’
preferences in palliative care: a systematic mixed studies review. Palliative
Medicine, 29(5), 399-419.
Seal, K., Murray, C., & Seddon, L. (2015). Family stories of end-of-life cancer care
when unable to fulfil a loved one’s wish to die at home. Palliative and Supportive
Care, 13(3), 473-83.
Shamseer, L., Moher, D., Clarke, M., Ghersi, D., et al. (2015). PRISMA-P Group.
Preferred reporting items for systematic review and meta-analysis protocols
(PRISMA-P) 2015: elaboration and explanation. British Medical Journal, 349:g7647,
viewed May 2016. www.prisma-statement.org.
Shih, C., Hu, W., Cheng, S., Yao, C., Chen, C., Lin, Y., & Chiu, T. (2015). Patient
preferences versus family physicians’ perceptions regarding the place of end-of-life
care and death: A nationwide study in Taiwan. Journal of Palliative Medicine, 18(7),
625-630.
Singer, P., Martin, D., & Kelner, M. (1999). Quality end-of-life care patient
perspectives. Journal of the American Medical Association, 281(2), 163-168.
Smyth, D., Farnell, A., Dutu, G., Lillis, S., et al. (2010). Palliative care provision by
rural general practitioners in New Zealand. Journal of Palliative Medicine, 13(3),
247-250.
Sofaer, S. (1999). Qualitative methods: what are they and why use them? Health
Services Research, 34(5), 1101-1118.
Spice, R., Paul, L.R. & Biondo, P.D. (2012). Development of a rural palliative care
program in the Calgary zone of Alberta Health services. Journal of Pain and
Symptom Management, 43(5), 911-924.
Steinhauser, K., Christakis, N., Clipp, E., et al. (2000). Factors considered important at
the end of life by patients, family, physicians and other care providers. Journal of the
American Medical Association, 284(19), 2476-2482.
Swerissen, H., & Duckett, S. (2014). Dying Well. Grattan Institute Report 2014-10,
viewed 24 November 2016. http://grattan.edu.au/wp-content/uploads/2014/09/815-
dying-well.pdf
Tamannai, M., Kaah, J., &Mbah, G. (2015). An evaluation of a palliative care outreach
programme for children with Burkitt lymphoma in rural Cameroon. International
Journal of Palliative Nursing, 21(7), 331-337.
250
Tang, S., Wu, S., Hung, Y., et al. (2009). Trends in quality of end-of-life care for
Taiwanese cancer patients who died in 2000-2006. Annals of Oncology, 20, 343–
348.
Thomas, C., Morris, S., & Gatrell, T. (2003). Place of death in the Morecambe Bay
area: patterns of and preferences for place of final care and death among terminally
ill cancer patients and their carer. Final report to The Research and Development
Department, NHS Directorate of Health and Social Care North. Reference
RDO/28/3/10; January 2003.
Thomas, C., Morris, S.M., & Clark, D. (2004). Place of death: preferences among
cancer patients and their carers. Social Science and Medicine, 58, 2431-2444.
Tricco, A.C., Lille, E., Zarin, W., et al. (2016). A scoping review on the conduct and
reporting of scoping reviews. BMC Medical Research Methodology, 16:15, doi:
10.1186/s12874-016-0116-4 [online].
Ulin, P., Robinson, E., & Tolley, E. (2005). Collecting qualitative data: the science and
the art. Qualitative methods in public health: a field guide for applied research.
Family health international, San Francisco: Jossey Bass, p. 71-104.
van der Geest, S. (2004). Dying peacefully: considering good death and bad death in
Kwahu-Tafo, Ghana. Social Science and Medicine, 58, 899-911.
Veillette, A., Fillion, L., Wilson, D., et al. (2010). La belle morte en milieu rural: a
report of an ethnographic study of the good death for Quebec rural francophones.
Journal of Palliative Care, 26(3), 159-166.
Vincent, C., Burnett, S., Carthey, J. (2014). Safety measurement and monitoring in
healthcare: a framework to guide clinical teams and healthcare organisations in
maintaining safety. BMJ Quality and Safety, 23, 670-677.
Walters, G. (2004). Is there such a thing as a good death? Palliative Medicine, 18, 404-
408.
Welte, T., Shield, R., Teno, J., Miller S., & Welch, L. (2005). Family perspectives on
end-of-life care experiences in nursing homes. The Gerontologist, 45(5), 642-650.
White, C., & Hardy, J. (2010). What do palliative care patients and their relatives think
about research in palliative care? - a systematic review. Supportive Care and Cancer,
18, 905-911.
Wiik, G.B. (2011). Don’t become a burden and don’t complain: a case study of older
persons suffering from incurable cancer and living alone in rural areas. Nursing
Reports, 1(e3), 7-14.
Williams, A., Duggleby, W., Eby, J., et al. (2013). Hope against hope: Exploring the
hopes and challenges of rural female caregivers of persons with advanced cancer.
BMC Palliative Care, 12(1): [online].
251
Wilson, D.M., Justice, C., Sheps, S., Thomas, R., Reid, P., & Leibovici, K. (2006).
Planning and providing end-of-life care in rural areas. National Rural Health
Association, 22(2), 174-181.
Wilson, D.M., Fillion, L., Thomas, R., Justice, C., Bhardwaj, P.P., & Veillette, A.
(2009a). The "good" rural death: a report of an ethnographic study in Alberta,
Canada. Journal of Palliative Care, 25(1), 21-29.
Wilson, D.M., Fillion, L., Thomas, R., Justice, C., Veillette, A., & Bhardwaj P.P.
(2009b). Planning and providing for a good death using rural French-Canadian and
English-Canadian insights. Reviews in the Neurosciences, 20, 313-319.
Wilson, D.M., Truman, C.D., Thomas, R., et al. (2009c). The rapidly changing location
of death in Canada, 1994-2004. Social Science and Medicine, 68(10), 1752-1758.
Wohleber, A., McKitrick, D., & Davis, S. (2012). Designing research with hospice and
palliative care populations. American Journal of Hospice and Palliative Care, 29(5),
335-345.
World Health Organisation. (2002). National cancer control programmes: policies and
managerial guidelines, 2nd ed., report, Geneva, 1 April, 2017.
http://www.who.int/cancer/publications/nccp2002/en/
World Health Organisation. Palliative Care: the solid facts 2004. World Health
Organization Europe; viewed 11 October 2014.
www.euro.who.int/en/publications/abstracts/palliative-care.-the-solid-facts.
World Health Organization. (2015). Definition of palliative care, viewed 28 November
2015. http://www.who.int/cancer/palliative/definition/en/.
Yamagishi, A., Morita, T., & Miyashita, M. (2012). Preferred place of care and place of
death of the general public and cancer patients in Japan. Supportive Care and Cancer,
20, 2575-2582.
Yasumura, S., Haga, H., Shibata, H., et al. (2000). Factors relating to place of death of
Japanese people from a small town in a rural area. Aging Clinical and Experimental
Research, 12(6), 449-454.
252
253
APPENDICES
Appendix 1: HREC ethics approvals
Appendix 2: Quality assessment score cards for the scoping and systematic literature
reviews
Appendix 3: Numerical summary of the studies included in the ‘good death’ scoping
review
Appendix 4: Participant information sheets and consent form
Appendix 5: Interview prompts
Appendix 6: Copies of published papers
254
Appendix 1: HREC ethics approvals.
255
256
257
258
259
260
261
Appendix 2: Quality assessment score cards for the scoping and
systematic literature reviews (Hawker et al. (2002); Gomes et al. (2013) and
further modified by the author).
1. Quality assessment for rural ’good death’ scoping review Score
Poor=0
Fair=1
Good=2
Title & abstract
Clear description of study relevant to rural ’good death’
Good background and aims relating to rural concept of the ’good death’
Appropriate method, clearly explained
Sample strategy appropriate to aims/ sample size/response rate
Rigorous analysis
Credible (>1 assessor, peer review; quotes; time)
Ethics addressed
Reflexivity and aware of own bias
Clear statement of findings
Results generalizable? Transferable to wider rural population?
Importance to policy and practice
Rural definition
Informant: Prospective patient/carer; or HCP currently working =2
Retrospective >6months=1
Well community =0
Total /24, >70% high quality
High ≥17; med ≥10-16; Low <10
2. Quality assessment for rural PoD (actual) systematic review (quantitative)
Score
Poor=0
Fair=1
Good=2
Design : prospective =2; cross-sectional=1; retrospective=0
Sufficient description of rural sample
Adjustments made for confounders
Representative sample/relevant population
Explicit inclusion/exclusion criteria
Rural response rate: >60%=2; 59-50%=1; <50%=0
Rural variable reliable and objective/reliable and objective assessment tool
Rigorous. appropriate analysis
Total /16, >75% high quality
High ≥12; med ≥;8-11; low <8
262
3. Quality assessment for rural PoD systematic review (preferred place) Score
Poor=0
Fair=1;
Good=2
Title & abstract
Clear description of study relevant to rural preferred PoD.
Good background and aims relating to rural preferred PoD.
Appropriate method, clearly explained
Sample strategy appropriate to aims/ sample size/response rate
Rigorous analysis
Credible (>1 assessor, peer review; quotes; time)
Ethics addressed
Reflexivity and aware of own bias
Clear statement of findings
Results generalizable? Transferable to wider rural population?
Importance to policy and practice in improving rural PoD
Total /20, >75% high quality
High ≥15; med ≥10-14; Low <10
3.4 Quality assessment for rural end-of-care from perspectives of rural
patients and FCGs Score
Poor=0
Fair=1
Good=2
Title & abstract
Clear description of study relevant to rural palliative patient/FCG
perspectives
Good background and aims relating to rural palliative patient/ FCG
perspectives
Appropriate method, clearly explained
Sample strategy appropriate to aims/ sample size/response rate
Rigorous analysis
Credible (>1 assessor, peer review; quotes; time
Ethics addressed
Reflexivity and aware of own bias
Clear statement of findings
Results generalizable? Transferable to wider rural population?
Importance to policy and practice in improving rural PC by listening to
patient/FCG perspectives
Rural definition
Informant: PT/FCG perspective
Prospective: Patient/carer =2
Retrospective >6months=1
2nd hand perspectives
Total /24, >75% high quality
High ≥18; med ≥13-17; Low <13
263
Appendix 3: Numerical summary of the studies included in the ‘good
death’ scoping review (Chapter 4; Section 4.2.1)
Developed countries
(n=10)
Developing countries
(n=8)
Date of publication 2000-2004 1 3 2005-2009 3 2 2010-2015 6 3
Countries Developed (n=10)
North America 7 0 United Kingdom 2 0
Norway 1 0 Developing (n=7) Africa 0 5
Bangladesh 0 1 Vietnam 0 1 Papua New Guinea
0 1
Source of article Journal 9 8 Dissertation 1 0
Source discipline Palliative/hospice care 7 3 Social Science and Medicine
0 2
Other nursing 2 0 Population Health 0 1 Anthropology 0 1 Cross cultural gerontology 0 1 Theology 1 0
Study objective included “Good death”
4 3
Type of article
Qualitative (n=17) Phenomenological 3 0 Ethnographic 2 1 Open ended written survey
2 0
Anthropological field work
0 3
Not specified 3 3 Systematic review 0 1
Methods* Interviews 7 4 Focus groups 5 3 Observations/Field work 0 4
Written surveys 2 0 Clinical observations 0 3
Terminology Good death 7 4 Peaceful 0 2 Other 3 1 Bad death 0 1
264
Appendix 4: Participant information sheets and consent
PATIENT PARTICIPANT INFORMATION SHEET
Principal Researcher: Dr Sue Rainsford (Sharp Street Surgery, Cooma and PhD student ANU) Associate Researchers: Dr Rob Wiles (Sharp Street Surgery, ANU Rural Medical School, Cooma) Prof Nick Glasgow (ANU Medical School) A/Prof Christine Phillips (ANU Medical School) Prof Rod MacLeod (Hammond Care and Sydney Uni. Medical School)
A RURAL PERSPECTIVE ON QUALITY END-OF-LIFE CARE
You are invited by Sue Rainsford and the associate research team to participate in a study that will form the basis of her PhD research thesis. Before you decide whether or not to participate in this study, it is important for you to understand why the research is being done and what it will involve. Please take time to read the following information carefully and discuss it with others if you wish. Description and Methodology: The study will explore what factors rural people (patients, families, unofficial carers, general practitioners and palliative care/community nurses) see as important in providing and receiving quality care at the end of a person’s life and the advantages and disadvantages of living in a rural area. The study will focus on the priority given by patients and their families to receive care and die in their preferred place. Data will be collected through face-to-face interviews. Participants: The study will be carried out in south eastern NSW in the 3 local government areas of Cooma Monaro, Snowy River and Bombala. It is hoped to recruit 20 patients living in Cooma, smaller townships and villages, and living on isolated farms. Participants will be patients (and a family member or unofficial carer and their GP and/or palliative care nurse) diagnosed with a life limiting illness and a life expectancy of less than 12 months. Use of Data and Feedback: The results of this study will be written up in a thesis to be read by examiners and a copy stored in the ANU library. Articles may also be submitted for publication in Medical Journals or presented at national or international conferences. Voluntary Participation & Withdrawal: Participation in this study is voluntary. It is completely up to you whether or not you participate. If you decide not to participate, it will not affect the treatment or care you receive now or in the future.
If you wish to withdraw from the study once it has started, you can do so at any time, until the work is prepared for publication, without having to give a reason. If you withdraw from the
study, where possible your data including personal quotes will be destroyed however, it may not be possible to destroy all your data from the study results particularly if your identifying details have been removed. With your permission individual quotes may be used in the written report. What does participation in the research request of you? If you agree to participate in this study Sue Rainsford will contact you to answer any questions, obtain consent and arrange a convenient time and place to visit. The study will involve a 1 hour audiotaped interview. This may be shorter if you are not feeling well. You will be asked questions about your illness, your care, your needs and support (what’s good, what could be done better), where you would prefer to be cared for in the last weeks and days and if you have a preference for where you die. After the interview the audiotape will be transcribed. The aim is to discover common and unique threads relating to end of life care in this region. Where possible a follow up interview will be carried out at a convenient time.
265
For those who would like to participate, Sue Rainsford will also like to talk to your husband/wife/partner or main carer to explore their views on their care needs and support and their preferences regarding your place of care and death. Ideally this will be carried out at a separate time but if you prefer can be done at same time as your interview. The interviews will be audiotaped, transcribed and discussed looking for common and unique themes within the household and the community. Sue Rainsford will also contact this person about 2-3 months after you have died to gain a retrospective view on the care and support you both received. With your permission Sue will also talk to your GP or palliative care/community nurse to talk about their perspective on your illness and care, and the issues they faced trying to achieve your preferred place of care and death. Again these interviews may be audiotaped, transcribed, coded for themes with results compared to those received from their patient and carer and between health professionals. If you are interested, a few patients will be invited to be the subject of a case study. This will require a number of interviews over the course of your illness. The purpose is to look more deeply at what it is really like for a rural person to have a life limiting illness, the day to day struggles and blessings faced, how you make sense of your illness, how you adjust to the changes in plans and preferences and what influence rural living has on your care. If you wish to participate you will be asked to sign a consent form: this will be clearly explained to you by the researcher. Location and Duration: Interviews will be conducted at your home or another mutually convenient location and last up to 1 hour. For most participants the total time requested of you in this research is 2 hours. For those willing to participate in a case study the total time will be 6-8hours over a number of months. Risks: If you decide to participate in the study you need to be aware that talking about your illness and life expectancy can be distressing. The research team has taken steps to minimize any risk to you. Sue Rainsford will be conscious of the sensitive nature of the interviews and your frail physical condition. The interview can be stopped at any time. If you suffer any harm or complications as a result of the study, you should contact the researcher as soon as possible, who will assist you in obtaining appropriate care or counselling. With your permission your GP, nurse or other appropriate person identified by you will be informed of any significant distress you experience. If you require immediate counselling please phone Beyond Blue on 1300 224 636. Every attempt will be made to ensure information is provided in such a way that you cannot be identified however it is possible you may be recognised by your stories or quotes. To minimise this risk neither your name nor any other identifying information will be published and your age will be changed in any written report. Benefits: The study aims to further our knowledge about rural end-of-life care and preferences of place of care and place of death. Although this study is unlikely to benefit you directly, benefits may be seen in the future through improvements in rural palliative care. Confidentiality: Any identifiable information that is collected about you in connection with this study will remain confidential and be disclosed only with your permission, or except as required by law. Only the researchers named above will have access to your details and results. Confidentiality of the information collected in connection with this study will be maintained at all times. Consent forms and researcher personal notes will be stored in a secure location. All study data will be de-identified and stored on a password-protected computer. Five years after publication the data will be disposed of by shredding and erasure of computer-generated data Contact Details for More Information: As mentioned by your GP or palliative care nurse, the primary researcher, Sue Rainsford, will phone you in about a week to discuss the study with you and to answer any queries you may have. If you would like to know more at any stage, please either contact Sue at Sharp Street Surgery, 6455 0000 or email suzanne.rainsford@anu.edu.au or Nicholas Glasgow (supervisor) nicholas.glasgow@anu.edu.au
266
The ethical aspects of this research have been approved by both ANU and GW Human Research Ethics Committees. If you have any concerns or complaints about how this research has been conducted, please contact ANU Human Research Ethics Committee. Contact the Ethics Manager, Australian National University, Telephone +61 2 61253427; email: Human.Ethics.Officer@anu.edu.au Quote 2014/736
Greater Western Human Research Ethics Committee (HREC) of the Western NSW & Far West
Local Health Districts. Contact The Executive Officer, Greater Western Human Research Ethics
Committee, Western NSW Local Health District. PO Box 143 Bathurst NSW 2795 or telephone
(02) 6330 5941. Quote: HREC/14/GWAHS/122
The Chief Executive’s delegate for the Southern NSW Local Health District (LHD) has authorized this study within the Southern NSW LHD.
Thank you for taking the time to consider this study. If you wish to take part please sign the attached consent form and either send it to: Dr S Rainsford, Sharp Street Surgery, PO Box 159, Cooma 2630 or have ready for collection at the interview. This information sheet is for you to keep.
267
FAMILY/ CARER PARTICIPANT INFORMATION SHEET
Principal Researcher: Dr Sue Rainsford (Sharp Street Surgery, Cooma and PhD student ANU) Associate Researchers: Dr Rob Wiles (Sharp Street Surgery, ANU Rural Medical School, Cooma) Prof Nick Glasgow (ANU Medical School) A/Prof Christine Phillips (ANU Medical School) Prof Rod MacLeod (Hammond Care and Sydney Uni. Medical School)
You are invited by Sue Rainsford and the associate research team to participate in a study that will form the basis of her PhD research thesis. Before you decide whether or not to participate in this study, it is important for you to understand why the research is being done and what it will involve. Please take time to read the following information carefully and discuss it with others if you wish. Description and Methodology: The study will explore what factors rural people (patients, families, unofficial carers, general practitioners and palliative care/community/aged care nurses) see as important in providing and receiving quality care at the end of a person’s life and the advantages and disadvantages of living in a rural area. The study will focus on the priority given by patients and their families to receive care and die in their preferred place. Data will be collected through face-to-face interviews. Participants: The study will be carried out in south eastern NSW in the 3 local government areas of Cooma Monaro, Snowy River and Bombala. It is hoped to recruit 20 families or unofficial carers of patients living in Cooma, smaller townships and villages, and living on isolated farms. Participants will be patients (and a family member or unofficial carer and their GP and/or palliative care/aged care nurse) diagnosed with a life limiting illness and a life expectancy of less than 12 months. Use of Data and Feedback: The results of this study will be written up in a thesis to be read by examiners and a copy stored in the ANU library. Articles may also be submitted for publication in Medical Journals or presented at national or international conferences. Voluntary Participation & Withdrawal: Participation in this study is voluntary. It is completely up to you whether or not you participate. If you decide not to participate, it will not affect the treatment or care your family member/patient receives now or in the future.
If you wish to withdraw from the study once it has started, you can do so at any time, until the work is prepared for publication, without having to give a reason. If you withdraw from the study, where possible your data including personal quotes will be destroyed; however, it may not be possible to destroy all your data from the study results particularly if your identifying details have been removed. With your permission individual quotes may be used in the written report. What does participation in the research request of you? If you agree to participate in this study Sue Rainsford will contact you to answer any questions, obtain consent and arrange a convenient time and place to visit. The study will involve a 1 hour audiotaped interview. You will be asked questions about your needs and support in caring for your family member or patient (what’s good, what could be done better), and your preferences regarding the patient’s place of care and death. Ideally this will be carried out at a separate time to the patient’s interview. After the interview the audiotape will be transcribed. The aim is to discover common and unique threads relating to end of life care in this region. Where possible a follow up interview will be carried out at a convenient time. Sue Rainsford will also contact you about 2-3 months after your family member/ patient has died to gain a retrospective view on the care and support you both received. Sue will be also be interviewing your family member/patient, at a separate time, to ask questions about their illness, their care, their needs and support (what’s good, what could be done better),
268
where they would prefer to be cared for in the last weeks and days and if they have a preference for where they die. If for any reason your family member/ patient decides not to participate, with their permission you can participate to discuss your views as outlined above. Likewise, if you decide not to participate the patient can still be involved in the study as we will be discussing their illness and preferences. A few patients will be invited to be the subject of a case study. This will require a number of interviews over the course of their illness. The purpose is to look more deeply at what it is really like for a rural person to have a life limiting illness, the day to day struggles and blessings faced, how they make sense of their illness, how they adjust to the changes in plans and preferences and what influence rural living has on their care. If you agree you may be included in this longitudinal study. If you wish to participate you will be asked to sign a consent form: this will be clearly explained to you by the researcher. Location and Duration: Interviews will be conducted at your home or another mutually convenient location and last up to 1 hour. For most participants the total time requested of you in this research is 2 hours. For those willing to participate in a case study the total time will be 6-8hours over a number of months. Risks: If you decide to participate in the study you need to be aware that talking about illness and life expectancy can be distressing. The research team has taken steps to minimize any risk to you. Sue Rainsford will be conscious of the sensitive nature of the interviews. The interview can be stopped at any time. If you suffer any harm or complications as a result of the study, you should contact the researcher as soon as possible, who will assist you in obtaining appropriate care or counselling. With your permission your GP, palliative care nurse or other appropriate person nominated by you will be informed of any significant distress you experience. If you require immediate counselling please phone Beyond Blue on 1300 224 636. Every attempt will be made to ensure information is provided in such a way that you cannot be identified however it is possible you may be recognised by your stories or quotes. To minimise this risk neither your name nor any other identifying information will be published and your age and family member’s age will be changed in any written report. Benefits: The study aims to further our knowledge about rural end-of-life care and preferences of place of care and place of death. Although this study is unlikely to benefit you directly, benefits may be seen in the future through improvements in rural palliative care. Confidentiality: Any identifiable information that is collected about you in connection with this study will remain confidential and be disclosed only with your permission, or except as required by law. Only the researchers named above will have access to your details and results. Confidentiality of the information collected in connection with this study will be maintained at all times. Consent forms and researcher personal notes will be stored in a secure location. All study data will be de-identified and stored on a password-protected computer. Five years after publication the data will be disposed of by shredding and erasure of computer-generated data Contact Details for More Information: As mentioned by your family member’s GP or palliative care nurse, the primary researcher, Sue Rainsford, will phone you in about a week to discuss the study with you and to answer any queries you may have. If you would like to know more at any stage, please either contact Sue at Sharp Street Surgery, 6455 0000 or email suzanne.rainsford@anu.edu.au or Nicholas Glasgow (supervisor) nicholas.glasgow@anu.edu.au The ethical aspects of this research have been approved by both ANU and GW Human Research Ethics Committees. If you have any concerns or complaints about how this research has been conducted, please contact
269
1. ANU Human Research Ethics Committee. Contact the Ethics Manager, Australian National University, Telephone +61 2 61253427; email: Human.Ethics.Officer@anu.edu.au Quote 2014/736
2. Greater Western Human Research Ethics Committee (HREC) of the Western NSW &
Far West Local Health Districts. Contact The Executive Officer, Greater Western
Human Research Ethics Committee, Western NSW Local Health District. PO Box 143
Bathurst NSW 2795 or telephone (02) 6330 5941. Quote: HREC/14/GWAHS/122
The Chief Executive’s delegate for the Southern NSW Local Health District (LHD) has authorized this study within the Southern NSW LHD.
Thank you for taking the time to consider this study. If you wish to take part please sign the attached consent form and either send it to: Dr S Rainsford, Sharp Street Surgery, PO Box 159, Cooma 2630 or have ready for collection at the interview. This information sheet is for you to keep.
270
HEALTH PROVIDER PARTICIPANT INFORMATION SHEET
Principal Researcher: Dr Sue Rainsford (Sharp Street Surgery, Cooma and PhD student ANU) Associate Researchers: Dr Rob Wiles (Sharp Street Surgery, ANU Rural Med. School, Cooma) Prof Nick Glasgow (ANU Medical School) A/Prof Christine Phillips (ANU Medical School) Prof Rod MacLeod (Hammond Care and Sydney Uni. Medical School)
You are invited by Sue Rainsford and the associate research team to participate in a study that will form the basis of her PhD research thesis. Before you decide whether or not to participate in this study, it is important for you to understand why the research is being done and what it will involve. Please take time to read the following information carefully and discuss it with others if you wish. Description and Methodology: The study will explore what factors rural people (patients, families, unofficial carers, general practitioners and palliative care/community/aged care nurses) see as important in providing and receiving quality care at the end of a person’s life and the advantages and disadvantages of living in a rural area. The study will focus on the priority given by patients and their families to receive care and die in their preferred place. Data will be collected through face-to-face interviews. Participants: The study will be carried out in south eastern NSW in the 3 local government areas of Cooma Monaro, Snowy River and Bombala. It is hoped to recruit 20 patients living in Cooma, smaller townships and villages, and living on isolated farms. Participants will be patients (and a family member or unofficial carer and their GP and/or palliative care/aged care nurse) diagnosed with a life limiting illness and a life expectancy of less than 12 months. Use of Data and Feedback: The results of this study will be written up in a thesis to be read by examiners and a copy stored in the ANU library. Articles may also be submitted for publication in Medical Journals or presented at national or international conferences. Voluntary Participation & Withdrawal: Participation in this study is voluntary. It is completely up to you whether or not you participate. If you decide not to participate, it will not affect the excellent existing relationship between the local researchers and yourself and your practice.
If you wish to withdraw from the study once it has started, you can do so at any time, until the work is prepared for publication, without having to give a reason. If you withdraw from the study, where possible your data including personal quotes will be destroyed; however, it may not be possible to destroy all your data from the study results particularly if your identifying details have been removed. With your permission individual quotes may be used in the written report. What does participation in the research request of you? If you agree to participate you will be asked by the principal researcher, Sue Rainsford, to identify potential palliative care participants for the study. The inclusion criteria includes patients with a life limiting illness < 3-6 months life expectancy (longitudinal case studies <6-12 months); 18 years of age or over; English speaking; capable of providing informed consent; living in the targeted rural region. You will be required to tell the patient (and their carer) of the existence of the study and ask if they would consider taking part. If they would like to know more about the study you will provide them with a participant’s information sheet and consent form. You would then gain their verbal consent to pass on their name and phone number to the principal researcher. If they provide a verbal agreement you will need to pass on these details to Sue Rainsford. After the patient has died, and with their prior permission, you will advise the researcher of the death. You will be contacted by the researcher to talk about your perspective on their illness and care, and the issues you faced trying to achieve their preferred place of care and death. These interviews may be audiotaped, transcribed, coded for themes with results compared to those received from the patient and carer and between health professionals. If you wish to participate you will be asked to sign a consent form.
271
Location and Duration: Your initial time commitment for this study is dependent on the number of potential participants you are caring for. As you are only required to mention the study, gain verbal consent to provide their name and phone number to the researcher and forward this information by phone or email the time involved would be about 5 minutes. Post death interviews will be conducted at your surgery or another mutually convenient location and last up to 1 hour. Risks: If you decide to participate in the study you need to be aware that talking about your management of patients can be distressing. The interview can be stopped at any time. If you suffer any harm or complications as a result of the study, you should contact the researcher as soon as possible, who will assist you in obtaining appropriate care or counselling. If you require immediate counselling please phone Beyond Blue on 1300 224 636. Every attempt will be made to ensure information is provided in such a way that you cannot be identified; however, it is possible you may be recognised by your stories or quotes. To minimise this risk neither your name nor any other identifying information will be published and your age will be changed in any written report. As with any research, there may also be risks associated with the research that are presently unknown or unforeseeable. Benefits: The study aims to further our knowledge about rural end-of-life care and preferences of place of care and place of death. Although this study is unlikely to benefit you directly, benefits may be seen in the future through improvements in rural palliative care. Confidentiality: Any identifiable information that is collected about you in connection with this study will remain confidential and be disclosed only with your permission, or except as required by law. Only the researchers named above will have access to your details and results. Confidentiality of the information collected in connection with this study will be maintained at all times. Consent forms and researcher personal notes will be stored in a secure location. All study data will be de-identified and stored on a password-protected computer. Five years after publication the data will be disposed of by shredding and erasure of computer-generated data Contact Details for More Information: If you would like to know more at any stage, please either contact Sue Rainsford at Sharp Street Surgery, 6455 0000 or email suzanne.rainsford@anu.edu.au or Nicholas Glasgow (supervisor) nicholas.glasgow@anu.edu.au The ethical aspects of this research have been approved by both ANU and GW Human Research Ethics Committees. If you have any concerns or complaints about how this research has been conducted, please contact
1. ANU Human Research Ethics Committee. Contact the Ethics Manager, Australian National University, Telephone +61 2 61253427; email: Human.Ethics.Officer@anu.edu.au Quote 2014/736
2. Greater Western Human Research Ethics Committee (HREC) of the Western NSW &
Far West Local Health Districts. Contact The Executive Officer, Greater Western
Human Research Ethics Committee, Western NSW Local Health District. PO Box 143
Bathurst NSW 2795 or telephone (02) 6330 5941. Quote: HREC/14/GWAHS/122
The Chief Executive’s delegate for the Southern NSW Local Health District (LHD) has authorized this study within the Southern NSW LHD.
Thank you for taking the time to consider this study. If you wish to take part please sign the attached consent form and either send it to: Dr S Rainsford, Sharp Street Surgery, PO Box 159, Cooma 2630. This information sheet is for you to keep.
272
PLACE OF DEATH IN COOMA- MONARO, BOMBALA and THE SNOWY MOUNTAINS
PARTICIPANT (FUNERAL DIRECTOR) INFORMATION SHEET (V2 18/2/15)
Principal Researcher: Dr Sue Rainsford (PhD student, Medical School ANU). Associate Researchers: Dr Rob Wiles (ANU Rural Medical School, Cooma) Prof Nick Glasgow (ANU Medical School) A/Prof Christine Phillips (ANU Medical School) Prof Rod MacLeod (Hammond Care) (Please note this research project is unrelated to Drs Rainsford and Wiles’ association with the Sharp Street Surgery and Cooma District Hospital)
Description and Methodology: This study aims to describe the actual place of death of residents living in the 3 local government areas of Cooma-Monaro, Bombala and Snowy River. De-identified socio-demographic data including place of death (home, local hospital, other hospital, ACT hospice, residential aged care facility, other) and last place of residence (post code only plus an indication if town, village or farm) and if burial or cremation will be retrieved from all death records held by the one local funeral director (FD) between January 2015 and December 2016 and emailed to the principal investigator at the end of each month. Data will be statistically analysed to determine how many residents in this region die at home. Participant: Local funeral director (FD). Use of Data and Feedback: The results of this study will be written up in a thesis to be read by examiners and a copy will be stored in the ANU library. Articles may also be submitted for publication in Medical Journals, presented at national or international conferences or used for teaching purposes. Your name will not be published. A summary of the research will be provided to you at completion of the study. Voluntary Participation & Withdrawal: Participation in this project is voluntary and you may, without any penalty, decline to take part or withdraw from the research at any time until the work is prepared for publication. If you do withdraw every attempt will be made to destroy and not use data previously provided by you; however, this may not be possible as some data provided by you may have been amalgamated with data obtained from newspaper obituaries. At the time of withdrawing you will be given the option of allowing continued use of data supplied by you. What does participation in the research request of you? You will be asked to email to the researcher by the 5th day of each month de-identified data on local residents who have died in the previous month. To maintain anonymity and confidentiality it will be important you do not provide names or actual home addresses. You will be asked to provide socio-demographic data from your records including date of death, age at death, gender, place of death (home, local hospital, other hospital, ACT hospice, residential aged care facility, other), last place of residence (post code only plus an indication if town, village or farm), cause of death and if cremation or burial. If the data has not been received by the researcher by the 5th of each month the researcher will phone you. If data is still not received by the 12th day of the month the
273
researcher will phone you a second time. If data is still not received the researcher will not contact you until the 5th day of the next month. If data has not been received for 3 months and if the researcher has not been able to contact you by phone the researcher will send an email to confirm you have withdrawn from participation. No further contact will be made concerning the project. Location and Duration: The study is to be conducted in the 3 local government areas of Cooma-Monaro, Bombala and Snowy River between January 2015 and December 2016. (Once ethics approval has been granted you will be asked to backdate the first list to commence January 1 2015) Risks: While your name will not be identified in any report or presentation, as you are the only funeral director in the region it is possible your identity will be recognised by third-parties. Due to the difficulty of obtaining informed consent from relatives of deceased persons the researchers will provide you with a notice of intent to collect de-identified data to display in your office. The data provided by you will be combined with data obtained from newspaper obituaries. Confidentiality: Confidentiality will be protected as far as the law allows and your identity will be disclosed only with your permission, or except as required by law. Your name will not be used in published reports. Only the researchers named above will have access to the data you provide. Confidentiality of the information collected in connection with this study will be maintained at all times. Data Storage: Where: All study data will be stored on a password-protected computer and external hard drive. How long: Five years after publication the data will be disposed of by shredding and erasure of computer-generated data from hard drives and external storage devises. Contact Details for More Information: Any queries regarding the study should be directed to Sue Rainsford, suzanne.rainsford@anu.edu.au mobile: 0408485150 or Nicholas Glasgow (Chair of supervisory panel) nicholas.glasgow@anu.edu.au PH: 6125 2622 Ethics Committee Clearance: The ethical aspects of this research have been approved by the ANU Human Research Ethics Committee. If you have any concerns or complaints about how this research has been conducted, please contact: Ethics Manager ANU Human Research Ethics Committee Australian National University Telephone: +61 2 6125 3427 Email: Human.Ethics.Officer@anu.edu.au Approval 2015/005
274
275
Appendix 5: Interview prompt questions
Patient and carer interview prompts: Interviews will be conversational and semi structured without necessarily direct questioning. Sample prompts for both patient and carer (adapted from McCall & Rice 2005) include:
1. Can you tell me what you know about palliative care?
2. Can you tell me what palliative care services are available
locally?
3. If circumstances allowed you to choose, where do you think you
would prefer to be cared for? Why would you choose this place?
4. If circumstances allowed you to choose, where do you think you
would prefer to die? Why would you choose this place?
5. What do you think you will need to achieve this choice? To
receive the best care possible?
6. Under what circumstances do you think you may not be able to
achieve your choice?
7. Would you consider an alternative? Where would the alternative
be? Why?
8. How would you feel if you had to change your preference?
9. How important is it that you stay at home? (if that is what they
have indicated) For you? For your family?
10. Have you ever had to leave home to receive care? How did that
feel? Tell me what was good? What was bad?
11. Who helps you at home? (formal and informal carers)
12. What do think contributes to a ‘good death’?
13. Do you think you care is influenced by where you live?
positive/negative
276
277
Appendix 6: Copies of the published papers arising from this thesis
1. Rainsford, S., MacLeod, R.D., & Glasgow, N.J. (2016). Place of death in rural
palliative care: a systematic review. Palliative Medicine, 30(8), 745-63.
doi:10.1177/0269216316628779.
2. Rainsford, S., MacLeod, R.D., Glasgow, N.J., Wilson, D.M., Phillips, C.B., &
Wiles, R.B. (2016). Rural residents' perspectives on the rural 'good death': a scoping
review. Health and Social Care in the Community.
doi: 10.1111/hsc.12385. [Epub ahead of print, 19 September 2016].
3. Rainsford, S., MacLeod, R.D., Glasgow, N.J., Phillips, C.B., Wiles, R.B., &Wilson,
D.M. (2017). Rural end-of-life care from the experiences and perspectives of
patients and family caregivers: a systematic literature review. Palliative Medicine;
31(10); 895-912. doi: 10.1177/0269216316685234.
278
Palliative Medicine2016, Vol. 30(8) 745 –763© The Author(s) 2016Reprints and permissions: sagepub.co.uk/journalsPermissions.navDOI: 10.1177/0269216316628779pmj.sagepub.com
What is already known about the topic?
•• Most place-of-death studies have been population surveys or urban focused.•• Few studies have focused specifically on influencing factors or absolute importance of place of care and death in rural
settings.
What this paper adds?
•• Significant variations exist in rural definition, characteristics, available services and preferences.•• Quantitative studies indicate that rural-preferred place of death is home; however, qualitative studies highlight the complex-
ity and fluidity of decision-making that is not reflected in population surveys.•• Rural hospitals may act as substitute hospice.
Place of death in rural palliative care: A systematic review
Suzanne Rainsford1, Roderick D MacLeod2,3 and Nicholas J Glasgow1
AbstractBackground: There have been many studies on the actual and preferred place of care and death of palliative patients; however, most have been whole population surveys and/or urban focused. Data and preferences for terminally ill rural patients and their unofficial carers have not been systematically described.Aim: To describe the actual place of death and preferred place of care and/or death in rural palliative care settings.Method: A systematic mixed studies review using Preferred Reporting Items for Systematic Reviews and Meta-Analysis guidelines.Data source: PubMed, PsychINFO, Scopus and CINAHL databases were searched (September to December 2014); eligible quantitative and qualitative studies included preferred and/or actual place of death/care of rural, regional or remote residents; rural data that are clearly identifiable; death due to palliative condition (malignant and non-malignant) or survey of participants with current or hypothetical life-limiting illness.Results: A total of 25 studies described actual place of death; 12 preferred place of care or death (2 studies reported both); most deaths occurred in hospital with home as the preferred place of care/death; however qualitative studies suggest that preferences are not absolute; factors associated with place are not adequately described as rurality was an independent variable; significant heterogeneity (rural setting and participants), however, many areas had a greater chance of home death than in cities; rural data are embedded in population reports rather than from specific rural studies.Conclusion: Home is the preferred place of rural death; however, more work is needed to explore influencing factors, absolute importance of preferences and experience of providing and receiving palliative care in rural hospitals which often function as substitute hospice.
KeywordsPalliative care, place of death, patient preference, rural, systematic review, terminal care
1 Medical School, The Australian National University, Canberra, ACT, Australia
2HammondCare, Sydney, NSW, Australia3Palliative Medicine, The University of Sydney, Sydney, NSW, Australia
628779 PMJ0010.1177/0269216316628779Palliative MedicineRainsford et al.research-article2016
Review Article
Corresponding author:Suzanne Rainsford, Medical School, The Australian National University, C/- PO Box 159, Cooma, NSW 2630, Australia. Email: suzanne.rainsford@anu.edu.au
at Australian National University on August 11, 2016pmj.sagepub.comDownloaded from
746 Palliative Medicine 30(8)
Implications for practice, theory or policy
•• Further rural-specific studies are required to explore|| Factors influencing place of care and death.|| Significance of preferences.
•• There is a need to determine actual proportion of rural residents who die in their preferred place.
Introduction
Dying in one’s preferred place of death (PPOD) is one out-come defining quality palliative care1 and ‘the good death’ with the assumption that a home death is the ideal.2,3 The modern day ‘good death’ infers a ‘managed death’4,5 with the goal of not only adequate pain and symptom control but also autonomy and respect of patients’ and families’ wishes. A previous systematic review6 reported home to be the PPOD for 31%–87% of patients, 25%–64% of carers and 49%–70% of the general public.
Currently, most deaths occur in hospital7,8 with one multi-nation European study reporting hospital rates of up to 91% (Sweden)9 and home rates of between 45.4% (the Netherlands) and 12.8% (Norway).10 The median percent-age of all deaths in hospital is reported to be 54%.11 Some studies report a shift in deaths away from hospital12 towards home13 and care homes,14 while others report the opposite15–17 with a prediction that by 2030 home deaths in the United Kingdom will have decreased by 42% and insti-tutionalised deaths increased by 20%.18
However, a significant percentage of the world’s popula-tion lives outside a major capital city, for example, one-third of Australia’s population19 yet most studies reporting on preferences for place of care (POC) or place of death (POD) have been population surveys and/or urban focused. One of the difficulties in describing rural preferences is the lack of consensus in defining ‘rural.’ In general terms, ‘rural’ implies a geographical area outside cities and urban areas; however, ‘rural’ can mean different things to different peo-ple, both within and between nations. The degree of rurality can be based on population density, size of the community or remoteness as defined by proximity to services provided by capital cities or urban service centres. Since the early 1990s, three different geographic classifications have been developed in Australia: the Rural, Remote and Metropolitan Areas (RRMA) classification, the Accessibility/Remoteness Index of Australia (ARIA) and the Australian Standard Geographical Classification (ASGC).20 Depending on which classification is applied, it is possible for one small agricultural town of <2000 people, 160 km from the nearest major urban area, to be considered ‘rural’ based on popula-tion size (RRMA) or ‘inner regional’ based on the road dis-tance to accessible services (ASGC).
Despite the difficulty in defining ‘rural’, studies sug-gest that rural folk are different to their urban counter-parts21 and have unique perspectives and concerns about
dying in one’s home community22 with each rural location unique and different from the next. New models of care are being developed to expand community palliative care ser-vices, especially in rural and remote areas with one recom-mended outcome being an increase in home deaths;23,24 however, there is limited data describing the preferences of terminally ill rural patients and their unofficial carers. To our knowledge, no systematic reviews of international rural studies have been published. The aim of this review is to describe the actual place of death (APOD) and pre-ferred place of care (PPOC) and/or death in rural palliative care settings.
Method
This systematic review of both quantitative and qualitative studies was undertaken utilising the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guidelines.25
Literature search
Between September and December 2014, four electronic databases (PubMed, PsycINFO, Scopus and CINAHL) were searched using the following keywords and Medical Subject Heading (MeSH) terms: (choice OR prefer* OR decision) AND (palliative OR end-of-life OR terminal) AND (care OR death) AND (place OR location OR site) AND (Rural). A date filter was not applied, and all pub-lished international English language, peer reviewed research articles were identified.
The search was widened by limiting the keywords to (‘place of death’ OR ‘place of care’) AND (Rural) on the presumption that rural data may be embedded in popula-tion studies. The two searches were downloaded to ENDNOTE X7, merged and duplicates deleted. The refer-ence lists of all included studies were scanned for addi-tional articles. The researcher hand searched the recent issues (June 2013–April 2015) of six relevant journals (Palliative Medicine, Journal of Palliative Medicine, Journal of Pain and Symptom Management, Palliative and Supportive Care, Australian Journal of Rural Health and Journal of Rural Health). Relevant articles found inciden-tally or identified by PubMed monthly updates (January–May 2015) were also included. One corresponding author26
at Australian National University on August 11, 2016pmj.sagepub.comDownloaded from
Rainsford et al. 747
provided a copy of their completed report containing spe-cific rural data.
Selection criteria
Studies were assessed against predetermined inclusion cri-teria and included if they reported the preferred and/or actual POD/POC as either part of the research aim or find-ings; participants included rural, regional or remote residents (no standardised definition of rurality was used); rural data were clearly identifiable; cause of death included a pallia-tive condition (malignant and non-malignant) or survey included preferences if participants had current or hypo-thetical life-limiting illness; the article indicated that a research study had been conducted with method and results described. While this systematic review excluded paedi-atric palliative care, the age filter (⩾19 years) was not applied in the initial search as different cut-off ages are used, ranging from 15 to 19 years. Both qualitative and quantitative studies were included.
Rejected studies were either clearly irrelevant or those that addressed the topic in general but failed in one or more of the inclusion criteria. The titles of all identified studies were reviewed independently by two researchers (S.R. and N.J.G.). S.R. and R.D.M. independently read abstracts of studies identified by title and full transcripts where further clarification was required. Disagreements were discussed via email and resolved by consensus.
Data extraction
Data from each accepted study were extracted and entered into a table under the following headings: article informa-tion (author, year, country, aim); methods (design, setting, participants, rural definition, inclusion and exclusion crite-ria, rural response rate, assessment and interview question) and results. When provided, predictors of and reasons for preferences were recorded. Data were extracted by S.R. with a second reviewer (R.D.M.) independently assessing a random selection of six studies (17%).
Assessment of quality
The quality of each study was assessed on aspects relevant to rural POC/POD and was not necessarily an assessment of the study per se, resulting in some high-quality studies receiving a low score. All studies were rated to be of low, medium or high quality based on a simple scoring system described by Gomes et al.6 and modified by the researcher to account for the rural focus. Different criteria27 were used to assess the quantitative and qualitative28 studies with max-imum scores of 16 for quantitative studies and 20 for quali-tative. Both tools awarded a score of 2 (Good), 1 (Fair) and 0 (Poor). Items scored for quantitative studies included study design, adequate rural description, representative rural
population, explicit inclusion/exclusion criteria, rural response rate, reliable and objective rural variable, adjust-ments made for confounders and rigorous/appropriate anal-ysis. Qualitative studies were graded on clear title and abstract, clear aims and background, appropriate method and justification for using qualitative method, appropriate and representative sample, rigour in analysis and interpreta-tion, ethics considered, reflexivity and bias acknowledged, clear statement of findings, generalisability and importance of findings.
The quality assessment was made by S.R. with R.D.M. independently assessing, using the same scoring system, a random selection of six studies (17%). Differences were discussed by email and resolved by consensus.
Data synthesis
Studies were described in terms of country, date of publi-cation, populations (including palliative condition), design, rural description and assessment outcome. Two main categories were identified: APOD and PPOC/PPOD with results summarised in two separate tables. Due to het-erogeneity, a meta-analysis was not possible. Findings are reported descriptively.
Data were manipulated into a common format, and when possible algebraic back calculations were performed to determine the percentage of actual or preferred rural home deaths. Where the adjusted odds ratio (aOR) was reported, the rate of a rural home death was compared to urban where aOR of urban home death = 1. Themes regard-ing POC/POD are reported for qualitative studies.
Ethical approval was not required as this study was a systematic literature review.
Results
We identified 254 articles from the electronic searches. After merging searches and removing duplicates, 96 stud-ies were identified. Following review of titles, abstracts and full texts, where necessary, 23 studies were eligible for inclusion (Figure 1). Of those rejected, all included rural participants; however, five did not separate rural from urban data, seven did not relate to POC/POD and two reported data for all deaths including children and acci-dents. Hand-searches, review of reference lists and monthly PubMed updates (including May 2015) added 16 additional articles. A total of 39 papers10,13,29–65 were reviewed reporting 35 separate studies. One corresponding author supplied their full report26 with data merged with their published paper.62 Data extraction for six stud-ies29,37,39,55,60,65 required arithmetic calculations of the data by the researcher to enable results of rural POD to be pre-sented as a percentage. There was an initial disagreement on the quality score of two of the six studies selected for checking due to the interpretation of the extracted rural
at Australian National University on August 11, 2016pmj.sagepub.comDownloaded from
748 Palliative Medicine 30(8)
data. This was resolved by both the researchers reviewing the studies for the second time and applying strict views of the rural aspect resulting in agreement by consensus.
Characteristics of the studies
The studies, published from 1990 to 2015, reported rural APOD (n = 23, Table 1), PPOD (n = 6),37,40,48,56,60,63 PPOC (n = 3),47,64,65 PPOD and PPOC (n = 1)53 and both APOD and PPOD (n = 2).26,31,62 Data were collected from at least 278,687 rural residents (sample size ranged from 847 to 199,21138) from 21 countries (three studies did not pro-vide rural participant number32,57,59). Data were collected from Australia (7),39,41,48,52,54,56,59 Canada (6),13,36,40,42,51,63,64
England (4),10,26,38,57,58,62 Germany (2),35,49,55 Italy (2),10,34 Scotland (2),31,47 Taiwan (2),45,60 Japan (2),50,65 Spain (2)33,46 and one from each of Belgium,10,38 Botswana,43,44 China,37 Denmark,29 Korea,53 Mexico,32 Norway,10 the Netherlands,10,38 New Zealand,61 Sweden30 and Wales.10 One study10,38 col-lected data from more than one country.
A total of 6 studies reported national data,32,43–45,60,61,65 12 reported provincial, state or county data,13,29,30,33,35,36,41,42,51,
52,55,56,59,64 9 reported regional data,26,34,39,46,48,49,57,58,63 7 reported a single town or health service/practice,31,37,40,
47,50,53,54 and 1 reported that data set was multi-regional.10,38 Eight studies were described as rural or semi-rural,26,31,47,
48,50,57,58,61 one regional39 and the remainder were cross-sectional containing rural data. One study reported data
Figure 1. Flow diagram of selection process.Source: modified flow chart as described by Moher et al.25
POD: place of death; POC: place of care; APOD: actual place of death; PPOC/PPOD: preferred place of care/death.
at Australian National University on August 11, 2016pmj.sagepub.comDownloaded from
Rainsford et al. 749
Tab
le 1
. St
udie
s de
scri
bing
act
ual r
ural
pla
ce o
f dea
th.
Stud
y de
scri
ptio
nR
ural
N/t
otal
NSt
udy
desi
gnD
eced
ent
diag
nosi
sR
ural
des
crip
tion
Res
ults
Perc
enta
ge o
f ru
ral p
atie
nts
died
at
hom
e
Qua
lity
Aab
om e
t al
.29Fu
nen
coun
ty, D
enm
ark
Janu
ary
1996
–Dec
embe
r 19
98R
ural
N =
927
/438
6
Cro
ss-s
ectio
nal d
eath
ce
rtifi
cate
s lin
ked
with
fo
ur n
atio
nal a
nd t
wo
loca
l re
gist
ers.
POD
: hom
e or
inst
itutio
n (h
ospi
tal,
RA
CF)
Can
cer
Not
pro
vide
dR
ural
PO
Da :
hom
e =
29%
Urb
an h
ome
deat
hs =
30%
Influ
enci
ng fa
ctor
s: n
ot d
escr
ibed
sp
ecifi
cally
for
rura
l; ru
ral r
esid
ency
in
depe
nden
t va
riab
le.
29M
ediu
m
Axe
lsso
n an
d C
hris
tens
en30
Jam
tland
cou
nty,
Sw
eden
1990
Rur
al N
= 2
03
Cro
ss-s
ectio
nal d
eath
ce
rtifi
cate
s lin
ked
with
ho
spita
l rec
ords
.PO
D: h
ome,
hos
pita
l, R
AC
F
Dea
ths
from
GI
trac
t, br
east
, ur
ogen
ital,
skin
an
d th
yroi
d ca
ncer
s
Spar
sely
pop
ulat
ed
coun
tyR
ural
PO
D: h
ome
= 1
2%; N
H =
24%
, ho
spita
l = 6
4%In
fluen
cing
fact
ors:
pro
xim
ity t
o ho
spita
l and
sho
rt in
terv
al b
etw
een
diag
nosi
s an
d de
ath
^ ho
spita
l dea
ths
12H
igh
Burg
e an
d co
lleag
ues1
3,51
Nov
a Sc
otia
, Can
ada
1992
–199
7R
ural
N =
617
6/13
,652
Cro
ss-s
ectio
nal,
popu
latio
n-ba
sed
deat
h ce
rtifi
cate
s lin
ked
with
th
ree
adm
inis
trat
ive
heal
th
data
base
s.PO
D: h
ospi
tal o
r O
OH
(in
clud
ed b
oth
hom
e an
d N
H).
RR
R: 9
7%
Can
cer
Popu
latio
n <
1000
or
pop
ulat
ion
dens
ity <
400/
km2
Rur
al P
OD
: OO
H =
23.
8%,
hosp
ital =
76.
2%U
rban
OO
H d
eath
= 2
8.2%
Incr
easi
ng t
rend
for
OO
H d
eath
.In
fluen
cing
fact
ors:
not
des
crib
ed;
rura
l res
iden
cy in
depe
nden
t va
riab
le
23.8
Hig
h
Burn
s et
al.5
2
Sout
h A
ustr
alia
, Aus
tral
ia20
01–2
007
Rur
al N
= 2
205/
6849
Ran
dom
ised
, pop
ulat
ion-
base
d, fa
ce-t
o-fa
ce in
terv
iew
su
rvey
s w
ith b
erea
ved
rela
tives
; ann
ual s
tate
go
vern
men
t he
alth
sur
vey.
POD
: hom
e, h
ospi
tal,
hosp
ice,
RA
CF
Ter
min
al il
lnes
sPo
stco
de a
reas
ou
tsid
e th
e ca
pita
l ci
ty, >
1000
po
pula
tion
Rur
al P
OD
: hom
e =
18.
1%,
hosp
ital =
67.
5%, h
ospi
ce =
7.9
%,
RA
CF
= 6
.4%
Urb
an: h
ome
= 2
0%, h
ospi
tal =
55.
5%,
hosp
ice
= 1
6.2%
, RA
CF
= 8
.4%
(p
< 0
.001
)In
fluen
cing
fact
ors:
not
des
crib
ed;
rura
l res
iden
cy in
depe
nden
t va
riab
le
18.1
Hig
h
Car
roll3
1
Scot
land
1992
–199
4R
ural
N =
82
Ret
rosp
ectiv
e cr
oss-
sect
iona
l au
dit
of s
ingl
e pr
actic
e m
edic
al r
ecor
ds.
POD
: hom
e, c
omm
unity
ho
spita
l, re
side
ntia
l hom
e,
hosp
ital o
r ho
spic
e
Can
cer
Sem
i-rur
al
smal
l tow
n,
popu
latio
n =
11,
240
Rur
al P
OD
: hom
e =
30%
, com
mun
ity
hosp
ital =
17%
, res
iden
tial
care
= 1
0%, h
ospi
tal =
38%
ho
spic
e =
5%
Influ
enci
ng fa
ctor
s: n
ot r
epor
ted
30Lo
w
Cas
tillo
-Guz
man
et
al.32
Mex
ico
1999
–200
9N
= 5
,338
,322
dea
ths
Rur
al N
not
pro
vide
d
Cro
ss-s
ectio
nal,
popu
latio
n-ba
sed
Mex
ican
Nat
iona
l In
stitu
te o
f Sta
tistic
s an
d G
eogr
aphy
dat
abas
e.PO
D: h
ome,
hos
pita
l, ot
her
Can
cer
Not
pro
vide
dR
ural
PO
D: h
ome
= 6
9.4%
, ho
spita
l = 2
3.6%
, oth
er =
7%
Urb
an P
OD
: hom
e =
52.
3%,
hosp
ital =
43%
, oth
er =
4.7
%In
fluen
cing
fact
ors:
not
des
crib
ed;
rura
l res
iden
cy in
depe
nden
t va
riab
le
69.4
Med
ium (C
ontin
ued)
at Australian National University on August 11, 2016pmj.sagepub.comDownloaded from
750 Palliative Medicine 30(8)
Stud
y de
scri
ptio
nR
ural
N/t
otal
NSt
udy
desi
gnD
eced
ent
diag
nosi
sR
ural
des
crip
tion
Res
ults
Perc
enta
ge o
f ru
ral p
atie
nts
died
at
hom
e
Qua
lity
Cat
alan
-Fer
nand
ez e
t al
.33M
ajor
ca, S
pain
Apr
il 19
84–A
ugus
t 19
86R
ural
N =
135
/335
Ret
rosp
ectiv
e cr
oss-
sect
iona
l se
rvic
e-ba
sed
face
-to-
face
qu
estio
nnai
re s
urve
y w
ith
rela
tives
.PO
D: h
ome
or h
ospi
tal
Can
cer
‘Eve
ryw
here
on
the
isla
nd o
utsi
de t
he
capi
tal,
Palm
a de
M
allo
rca’
Rur
al P
OD
: hom
e =
65.
9%,
hosp
ital =
34.
1%U
rban
: hom
e =
31.
5%,
hosp
ital =
68.
5% (
p <
0.0
01)
Influ
enci
ng fa
ctor
s: n
ot d
escr
ibed
; ru
ral r
esid
ency
inde
pend
ent
vari
able
65.9
Med
ium
Coh
en e
t al
.10 a
nd
Hou
ttik
er e
t al
.38N
orw
ay, E
ngla
nd, W
ales
, Be
lgiu
m, I
taly
and
the
N
ethe
rlan
ds20
02, 2
003
Non
-met
ropo
litan
N =
199
,211
38/2
6,42
77
Cro
ss-s
ectio
nal C
omm
on
Euro
pean
dat
abas
e–de
ath
cert
ifica
tes.
POD
: hom
e ve
rsus
ou
tsid
e ho
me
(hos
pita
l, ca
re h
ome/
NH
), ot
her
(hos
pice
in E
ngla
nd a
nd t
he
Net
herl
ands
), pu
blic
roa
d,
wor
k
Life
lim
iting
illn
ess/
canc
erBa
sed
on p
opul
atio
n de
nsity
Rur
al P
OD
: hom
e –
Belg
ium
= 2
8.2%
, th
e N
ethe
rlan
ds =
39.
1%,
Engl
and
= 2
0.7%
(pa
lliat
ive
subs
et);
Can
cer
deat
hs: a
OR
hom
e de
ath
urba
n =
1Be
lgiu
m =
2.3
4, It
aly =
2.0
8, t
he
Net
herl
ands
= 1
.6, N
orw
ay =
1.6
6,
Engl
and
= 0
.9, W
ales
= 0
.71
Influ
enci
ng fa
ctor
s: n
ot d
escr
ibed
; ru
ral r
esid
ency
inde
pend
ent
vari
able
20.7
28.2
39.1
Hig
h
Cos
tant
ini e
t al
.34Li
guri
a an
d T
osca
na It
aly
1991
Rur
al N
= 1
664/
55,7
59
Ret
rosp
ectiv
e cr
oss-
sect
iona
l co
mpu
teri
sed
data
from
T
osca
na a
nd L
igur
ia R
egio
nal
Mor
talit
y R
egis
ters
,PO
D: h
ome
or h
ospi
tal
(incl
udin
g N
H)
Can
cer
Rur
al m
unic
ipal
ity
<15
0 in
habi
tant
s/km
2
Rur
al P
OD
: hom
e =
53.
4%;
Urb
an: 4
3.2%
; sem
i-urb
an: 6
0.4%
Tre
nd d
ecre
ase
in h
ome
deat
hs.
Influ
enci
ng fa
ctor
s: n
ot d
escr
ibed
; ru
ral r
esid
ency
inde
pend
ent
vari
able
53.4
Hig
h
Cra
wfo
rd54
Kan
garo
o Is
land
, Aus
tral
ia19
94–1
996
N =
15
Cro
ss-s
ectio
nal d
eath
ce
rtifi
cate
s w
ritt
en o
n K
anga
roo
Isla
nd.
POD
: hom
e, h
ospi
tal (
30
beds
)
Can
cer
Isla
nd p
opul
atio
n 40
00R
ural
PO
D: h
ome
= 2
7%,
hosp
ital =
73%
Influ
enci
ng fa
ctor
s: h
ospi
tal d
eath
s as
soci
ated
with
pro
xim
ity t
o ho
spita
l, sy
mpt
om m
anag
emen
t, ca
rer
need
s
26.7
Med
ium
Esco
bar
Pinz
on e
t al
.35,5
5
Rhi
nela
nd-P
alat
inat
e G
erm
any
Sept
embe
r 20
08–J
anua
ry
2009
Rur
al N
= 6
44/1
378
Cro
ss-s
ectio
nal w
ritt
en
surv
ey o
f rel
ativ
es; R
R: 3
6%PO
D: h
ome,
hos
pita
l, pa
lliat
ive
care
faci
lity
(hos
pice
, hos
pita
l, PC
U),
NH
, el
sew
here
(e.
g. w
ork,
on
holid
ays)
All
deat
hs (
72%
ha
d pr
ogre
ssiv
e di
seas
e).
25 M
ay–2
4 A
ugus
t 20
08
Rur
al c
lass
ifica
tion:
ru
ral m
unic
ipal
ity
<20
00 in
habi
tant
s,
rura
l tow
n 20
00–
<50
00, s
mal
l tow
n 50
00–<
20,0
00
Rur
al P
OD
a : ho
me
= 4
3%,
hosp
ital =
38%
, PC
U =
6%
, NH
= 1
1%,
else
whe
re =
2%
.La
rge
tow
n: h
ome
deat
h =
28.
4%In
fluen
cing
fact
ors:
not
des
crib
ed;
rura
l res
iden
cy in
depe
nden
t va
riab
le
43M
ediu
m
Tab
le 1
. (C
ontin
ued)
at Australian National University on August 11, 2016pmj.sagepub.comDownloaded from
Rainsford et al. 751
Stud
y de
scri
ptio
nR
ural
N/t
otal
NSt
udy
desi
gnD
eced
ent
diag
nosi
sR
ural
des
crip
tion
Res
ults
Perc
enta
ge o
f ru
ral p
atie
nts
died
at
hom
e
Qua
lity
Gat
rell
et a
l.57
Nor
th W
est,
Engl
and
1993
–200
0N
= 6
900
(rur
al n
ot
prov
ided
)
Cro
ss-s
ectio
nal;
linke
d m
orta
lity
data
pro
vide
d by
si
ngle
Hea
lth A
utho
rity
and
O
ffice
of N
atio
nal S
tatis
tics.
POD
: hom
e, h
ospi
tal,
hosp
ice,
els
ewhe
re, N
H/
retir
emen
t ho
me
Can
cer
Not
pro
vide
d;
incl
udes
sem
i-rur
al
Sout
h La
kela
nd
Res
iden
ts o
f sem
i-rur
al a
rea
are
mor
e lik
ely
to d
ie a
t ho
me
(aO
R =
1.3
34),
or h
ospi
tal
(aO
R =
1.6
77);
less
like
ly t
o di
e in
hos
pice
(aO
R =
0.6
40)
or N
H
(aO
R =
0.6
47)
than
urb
an (
aOR
= 1
).O
vera
ll PO
D: h
ospi
tal =
35%
, ho
spic
e =
33%
, hom
e =
22%
, no
plac
e of
dea
th r
ecor
ded
= 1
0%.
Influ
enci
ng fa
ctor
s: p
roxi
mity
to
hosp
ital o
r ho
spic
e in
crea
sed
rate
of
dyin
g th
ere
aOR
= 1
.334
95%
CI =
1.1
39–
1.56
2
Hig
h
Goo
drid
ge e
t al
.36Sa
skat
chew
an, C
anad
a20
04R
ural
N =
468
/109
8
Ret
rosp
ectiv
e co
hort
stu
dy
(CO
PD o
r lu
ng c
ance
r);
Link
ed d
ata
from
sev
en
sepa
rate
ser
vice
file
s in
clud
ing
vita
l sta
tistic
s de
ath
file.
POD
: hom
e, h
ospi
tal,
LTC
F
CO
PD o
r lu
ng
canc
erR
ural
cla
ssifi
catio
n:
rura
l/rem
ote
(⩽99
99 p
opul
atio
n si
ze),
smal
l urb
an o
r ur
ban
(⩾10
0,00
0)
Rur
al P
OD
: hom
e =
15.
4%,
hosp
ital =
59%
, LT
C =
25.
6%U
rban
hom
e de
aths
= 1
6.3%
.In
fluen
cing
fact
ors
of P
OD
wer
e no
t th
e fo
cus
of s
tudy
15.4
Med
ium
Her
d58
Wes
t C
umbr
ia, E
ngla
ndA
pril–
Oct
ober
198
7R
ural
N =
157
Cro
ss-s
ectio
nal,
sem
i-st
ruct
ured
inte
rvie
w
ques
tionn
aire
s w
ith la
y ca
rers
.PO
D: h
ome,
hos
pita
l
Can
cer
Sem
i-rur
al, s
mal
l to
wns
(m
axim
um
popu
latio
n =
28,
000)
, vi
llage
s, is
olat
ed
ham
lets
and
farm
s
Rur
al P
OD
: hom
e =
53%
Urb
an: h
ome
= 2
4–41
%In
fluen
cing
fact
ors:
dis
tanc
e fr
om
hosp
ital a
nd m
ore
likel
y to
die
at
hom
e; r
easo
ns fo
r ho
spita
l ad
mis
sion
: no
care
r (2
2%),
lay
care
r un
able
to
cont
inue
(45
%)
53H
igh
How
at e
t al
.39N
orth
Que
ensl
and,
A
ustr
alia
2004
N =
270
Ret
rosp
ectiv
e cr
oss-
sect
iona
l re
view
of m
edic
al c
hart
s;
sing
le p
allia
tive
care
ser
vice
POD
: hom
e, h
ospi
tal,
NH
Palli
ativ
e ca
reR
egio
nal u
sing
the
R
RM
A c
lass
ifica
tion
Rur
al P
OD
a : ho
me
= 1
1.6%
, ho
spita
l = 5
5.8%
, NH
= 0
.4%
, U
nkno
wn
= 2
4%R
emot
e PO
D: h
ome
= 0
, ho
spita
l = 3
5.3%
, NH
= 6
%,
unkn
own
= 5
8.9%
Met
ro: h
ome
= 2
3.4%
, ho
spita
l = 6
7.5%
, NH
= 6
.7%
, un
know
n =
2.4
%In
fluen
cing
fact
ors:
hom
e de
ath
– yo
unge
r, m
arri
ed, c
olor
ecta
l can
cer,
co
mm
unity
nur
se in
volv
emen
t
19M
ediu
m (Con
tinue
d)
Tab
le 1
. (C
ontin
ued)
at Australian National University on August 11, 2016pmj.sagepub.comDownloaded from
752 Palliative Medicine 30(8)
Stud
y de
scri
ptio
nR
ural
N/t
otal
NSt
udy
desi
gnD
eced
ent
diag
nosi
sR
ural
des
crip
tion
Res
ults
Perc
enta
ge o
f ru
ral p
atie
nts
died
at
hom
e
Qua
lity
Hun
t et
al.4
1
Sout
h A
ustr
alia
1990
–199
9R
ural
N =
752
7/29
,230
Popu
latio
n-ba
sed,
cro
ss-
sect
iona
l Sou
th A
ustr
alia
n ca
ncer
reg
istr
y.PO
D: h
ospi
tal,
hosp
ice,
ho
me,
NH
Can
cer
No
rura
l def
initi
onO
vera
ll PO
D: m
etro
polit
an p
ublic
ho
spita
l = 2
5.0%
, hos
pice
= 1
9.9%
, co
untr
y ho
spita
l = 1
6.9%
, pri
vate
re
side
nce
= 1
5.8%
; met
ropo
litan
pr
ivat
e ho
spita
l = 1
2.7%
; NH
= 9
.7%
; ci
ty (
low
soc
ioec
onom
ic)
hom
e de
ath
aOR
= 1
Influ
enci
ng fa
ctor
s: n
ot d
escr
ibed
; ru
ral r
esid
ency
inde
pend
ent
vari
able
aOR
= 3
.5 (
95%
C
I = 3
.03–
4.04
)H
igh
Jaya
ram
an a
nd Jo
seph
42
Briti
sh C
olum
bia,
Can
ada
2004
–200
8R
ural
N =
19,
428/
153,
111
Ret
rosp
ectiv
e po
pula
tion-
base
d de
ath
cert
ifica
tes.
POD
: hom
e, e
xten
ded
care
fa
cilit
y, h
ospi
tal o
r ot
her
All
deat
hs
(can
cer =
32%
, no
n-m
alig
nant
pr
ogre
ssiv
e di
seas
e =
33%
)
Post
code
Rur
al P
OD
: hom
e =
20.
2%, h
ospi
tal-
exte
nded
car
e =
20.
9%, o
ther
= 6
%V
anco
uver
: hom
e =
14.
6%, e
xten
ded
care
= 2
7.3%
, hos
pita
l = 5
5.8%
, ot
her =
2.4
%In
fluen
cing
fact
ors:
not
des
crib
ed;
rura
l res
iden
cy in
depe
nden
t va
riab
le =
52.
9%
20.2
Hig
h
Laze
nby
and
colle
ague
s43,
44
Bots
wan
aJu
ne 2
008
Rur
al N
= 1
0,48
7/18
,869
Rur
al o
ldes
t ol
d N
= 1
787
Ret
rosp
ectiv
e cr
oss-
sect
iona
l de
ath
cert
ifica
tes;
RR
: 97.
2%PO
D: h
ome
or h
ospi
tal
All
deat
hs
(HIV
/AID
S/ca
ncer
= 3
9%;
card
iac =
10%
; T
B =
4%
)
Rur
al c
ateg
orie
s as
iden
tifie
d by
th
e R
epub
lic o
f Bo
tsw
ana’
s C
entr
al
Stat
istic
s O
ffice
POD
: hom
e de
aths
rur
al (
38.3
%),
tow
n (2
9.7%
), ci
ty (
27.1
%).
Rur
al h
ome
deat
hs >
80 ye
ars:
50
%–1
00%
com
pare
d to
ove
rall
27%
–94%
.M
ost
hosp
ital d
eath
s oc
curr
ed in
ci
ties.
Influ
enci
ng fa
ctor
s: r
ural
hom
e de
ath
–fem
ale >
80 ye
ars
38.3
Med
ium
Lin
and
Lin4
5
Tai
wan
1997
–200
3R
ural
N =
21,
237/
18,8
69
Cro
ss-s
ectio
nal p
opul
atio
n-ba
sed
stud
y; d
eath
re
gist
ratio
n da
ta;
RR
: 96%
POD
: hom
e, h
ospi
tal
Can
cer
Nat
iona
l sta
ndar
dsR
ural
PO
D: U
rban
isat
ion
6 –
hom
e =
78.
4%, h
ospi
tal =
21.
6%,
Urb
anis
atio
n 7
– ho
me
= 8
1.7%
, ho
spita
l = 1
8.3%
Urb
anis
atio
n le
vel 1
(ci
ty):
hom
e =
36.
4%, h
ospi
tal =
63.
6%In
fluen
cing
fact
ors:
not
des
crib
ed;
rura
l res
iden
cy in
depe
nden
t va
riab
le
U6:
78.
4U
7: 8
1.7
Hig
h
Tab
le 1
. (C
ontin
ued)
at Australian National University on August 11, 2016pmj.sagepub.comDownloaded from
Rainsford et al. 753
Stud
y de
scri
ptio
nR
ural
N/t
otal
NSt
udy
desi
gnD
eced
ent
diag
nosi
sR
ural
des
crip
tion
Res
ults
Perc
enta
ge o
f ru
ral p
atie
nts
died
at
hom
e
Qua
lity
Lope
z-C
ampo
s et
al.4
6
And
alus
ia, S
pain
2010
and
201
1R
ural
N =
260
3/49
83
Ret
rosp
ectiv
e ob
serv
atio
nal
coho
rt s
tudy
; dea
th
cert
ifica
tes
POD
: hom
e, h
ospi
tal,
hosp
ice,
oth
er
CO
PDV
illag
es a
nd
coun
trys
ide.
17.
8%
of p
opul
atio
n
Ove
rall
POD
: C
OPD
– h
ome
= 3
5%,
hosp
ital =
57.
6%, h
ospi
ce =
6.7
%,
othe
r = 0
.7%
Non
-CO
PD: h
ome
= 3
3.8%
, ho
spita
l = 5
7.5%
, hos
pice
= 8
.2%
, ot
her
0.6%
aOR
CO
PD d
ying
at
hom
e:
urba
n =
1; r
ural
= aO
R 2
.8 (
95%
C
I = 2
.46–
3.16
).In
fluen
cing
fact
ors:
not
des
crib
ed;
rura
l res
iden
cy in
depe
nden
t va
riab
le
aOR
= 2
.8 (
95%
C
I = 2
.46–
3.16
)M
ediu
m
McN
amar
a59
Wes
tern
Aus
tral
ia,
Aus
tral
ia1
July
200
0–31
Dec
embe
r 20
02N
= 1
3,45
3N
o ru
ral N
Long
itudi
nal (
2.5
year
s)
popu
latio
n-ba
sed
retr
ospe
ctiv
e co
hort
des
ign;
de
ath
cert
ifica
tes
linke
d w
ith
hosp
ital a
nd p
allia
tive
care
nu
rsin
g re
cord
s.PO
D: u
sual
pla
ce o
f re
side
nce
(hom
e in
clud
ing
long
-ter
m r
esid
ents
of
RA
CF)
, hos
pita
l, ho
spic
e,
RA
CF
(tra
nsfe
rred
just
pri
or
to d
eath
) an
d ot
her
plac
e
PCS
Rur
al g
eogr
aphi
cal
rem
oten
ess
AR
IA
base
d on
acc
ess/
prox
imity
to
serv
ice
cent
res
Ove
rall
POD
(PC
S): u
sual
pla
ce
of r
esid
ence
= 3
2%, h
ospi
tal =
51%
, ho
spic
e =
11%
, RA
CF
= 3
%,
othe
r = 3
%; d
ying
in u
sual
pla
ce o
f re
side
nce
aOR
city
= 1
.(P
OD
of a
ll de
aths
: usu
al p
lace
of
resi
denc
e =
35.
8%, h
ospi
tal 4
8.6%
).In
fluen
cing
fact
ors:
not
des
crib
ed;
rura
l res
iden
cy in
depe
nden
t va
riab
le
Rur
al a
OR
= 0
.74
(95%
CI =
0.6
1–0.
91);
rem
ote
aOR
= 1
.03
(95%
C
I = 0
.75–
1.41
)
Med
ium
Papk
e an
d K
och4
9
Sach
sisc
he S
chw
eiz,
G
erm
any
1997
–200
3T
own
N =
131
5C
ount
ry N
= 1
001/
2316
Cro
ss-s
ectio
nal;
deat
h ce
rtifi
cate
s;PO
D: h
ospi
tal,
hom
e, N
H
Can
cer
Tow
n –
dist
rict
an
d la
rger
tow
ns;
coun
try
– sm
alle
r to
wns
and
vill
ages
aOR
tow
n ho
me
deat
h =
1; a
OR
co
untr
y ho
me
deat
h =
1.5
4 (9
5%
CI =
1.2
9–1.
82; p
= 0
).In
sm
all t
own
and
rura
l loc
atio
ns,
POD
from
ca
was
equ
ally
dis
trib
uted
be
twee
n ho
me
and
hosp
ital.
Influ
enci
ng fa
ctor
s: n
ot d
escr
ibed
; ru
ral r
esid
ency
inde
pend
ent
vari
able
aOR
= 1
.54
(95%
C
I = 1
.29–
1.82
)M
ediu
m (Con
tinue
d)
Tab
le 1
. (C
ontin
ued)
at Australian National University on August 11, 2016pmj.sagepub.comDownloaded from
754 Palliative Medicine 30(8)
Stud
y de
scri
ptio
nR
ural
N/t
otal
NSt
udy
desi
gnD
eced
ent
diag
nosi
sR
ural
des
crip
tion
Res
ults
Perc
enta
ge o
f ru
ral p
atie
nts
died
at
hom
e
Qua
lity
Smyt
h et
al.6
1
New
Zea
land
Rur
al N
= 1
86
Cro
ss-s
ectio
nal p
osta
l sur
vey
com
plet
ed b
y ru
ral G
Ps; R
R:
42.2
%;
POD
: hom
e, h
ospi
tal,
hosp
ice,
RA
CF
Palli
ativ
e ca
reG
P re
ceiv
ing
a ru
ral
GP
bonu
s id
entif
ied
by N
ew Z
eala
nd
Inst
itute
of R
ural
H
eath
Rur
al P
OD
: hom
e =
46.
5%,
NH
= 2
4.8%
, loc
al p
ublic
ho
spita
l = 1
6.7%
, hos
pice
= 7
.2%
; bas
e ho
spita
l = 4
.8%
.In
fluen
cing
fact
ors
for
not
dyin
g at
hom
e: c
areg
iver
bur
den
= 4
2%,
sym
ptom
s = 2
5%, p
atie
nt
choi
ce =
11%
, inv
estig
atio
ns =
5%
46.5
Low
Tho
mas
et
al.26
Nor
th W
est
Engl
and,
En
glan
dO
ctob
er 2
000–
Sept
embe
r 20
02R
ural
pat
ient
s N
= 1
2
Post
ber
eave
men
t in
terv
iew
s w
ith c
arer
or
heal
th
prof
essi
onal
.PO
D: h
ome,
hos
pita
l, ho
spic
e, e
lsew
here
Can
cer
Sem
i-rur
al r
egio
n of
Ken
dall/
Sout
h La
kela
nd
Rur
al P
OD
: hom
e =
25%
, ho
spic
e =
62.
5%, h
ospi
tal =
0,
else
whe
re =
12.
5% (
four
aliv
e at
en
d of
stu
dy: P
POD
hos
pice
= 3
, un
deci
ded
= 1
).In
fluen
cing
fact
ors:
not
rur
al s
peci
fic;
how
ever
, pro
xim
ity t
o ho
spic
e w
as
one
fact
or; r
ural
res
iden
cy o
ne
vari
able
2/8
= 2
5M
ediu
ma D
ata
obta
ined
fr
om fu
ll re
port
pr
ovid
ed b
y au
thor
; sm
all
com
pone
nt o
f a
larg
er s
tudy
Yas
umur
a et
al.5
0
Japa
nJa
nuar
y 19
86–D
ecem
ber
1988
Rur
al N
= 3
12
Ret
rosp
ectiv
e ru
ral s
tudy
; fa
ce-t
o-fa
ce in
terv
iew
s w
ith
care
r; R
RR
: 88.
6%PO
D: h
ome,
hos
pita
l
Dec
eden
t be
drid
den ⩾
7 da
ys
prio
r to
dea
th
Smal
l rur
al t
own;
po
pula
tion
13,0
00;
no h
ospi
tal o
r N
H,
no p
hysi
cian
s, o
r co
mm
unity
pal
liativ
e ca
re
Rur
al P
OD
: hom
e =
47.
7%
(Japa
nese
ave
rage
in 1
990
= 2
0%);
hosp
ital =
35.
9%.
Influ
enci
ng fa
ctor
s fo
r ho
me
deat
h:
age,
lack
of p
ain,
non
-can
cer,
sen
ility
, re
gula
r ho
me
visi
ts b
y a
publ
ic h
ealth
nu
rse,
livi
ng in
a t
hree
-gen
erat
iona
l ho
useh
old.
Fact
ors
hosp
ital d
eath
: can
cer,
hi
gher
edu
catio
n an
d liv
ing
with
sp
ouse
47.7
Hig
h
POD
: pla
ce o
f dea
th; O
OH
: out
of h
ospi
tal;
RA
CF:
res
iden
tial a
ged
care
faci
lity;
NH
: nur
sing
hom
e; L
TC
F: lo
ng-t
erm
car
e fa
cilit
y; R
RR
: rur
al r
espo
nse
rate
; RR
: ove
rall
resp
onse
rat
e; P
POD
: pre
ferr
ed
plac
e of
dea
th; G
P: g
ener
al p
ract
ition
er; G
I: ga
stro
inte
stin
al t
ract
; CI:
conf
iden
ce in
terv
al; C
OPD
: chr
onic
obs
truc
tive
pulm
onar
y di
seas
e; L
TC
: lon
g-te
rm c
are;
TB:
tub
ercu
losi
s; P
CS:
pal
liativ
e ca
re s
ubse
t; PC
U: p
allia
tive
care
uni
t; aO
R: a
djus
ted
odds
rat
io; R
RM
A: R
ural
, Rem
ote
and
Met
ropo
litan
Are
as.
a Dat
a re
calc
ulat
ed b
y re
sear
cher
to
enab
le r
ural
PO
D p
erce
ntag
es t
o be
rep
orte
d.
Tab
le 1
. (C
ontin
ued)
at Australian National University on August 11, 2016pmj.sagepub.comDownloaded from
Rainsford et al. 755
from non-cancer patients,46 eight both malignant and non-malignant,36,38,39,40,50,52,59,61 8 cross-sectional population data35,42–44,48,55,56,63–65 and the remainder reported data from terminally ill patients with cancer. In all, 31 studies were quantitative, 3 were qualitative47,48,63 and 1 was mixed.26,62
Heterogeneity
While 15 studies used death certificates to report the APOD, studies were heterogeneous in regard to rural set-ting, participants, inclusion criteria and assessment. While some studies listed a range of locations for POC/POD (home, hospital, hospice, nursing or care home), others restricted the options to ‘home or out of home/hospi-tal’10,29,33,34,38,43–45,50,58 or ‘hospital or out of hospital.’13,51 Long-term placement in a nursing home was considered ‘home’ for some studies13,51,59 and an ‘institution’ for oth-ers.29,34 Not all rural locations had the options of commu-nity hospital, inpatient hospice or nursing home. A total of 17 studies related to cancer patients and 8 to non-specified terminally ill patients. The definition of rural varied between the studies ranging from small remote, sparsely populated communities to large regional centres within large rural municipalities or provinces. Rural was described as isolated farms, hamlets, small towns or villages;31,46,49,58 low level of urbanisation;45 ‘everywhere on the island (Majorca) outside the capital’33 and sparsely populated rural provinces.30 One Australian study39 used the RRMA classification and another59 the ARIA. Rural populations ranged from <1000,49,51 <2000,35 4000,54 8000,40 <9999;36 rural towns 2000–<500035 and small towns <28,000.58 Some studies described as rural had access to large county hospitals (Sweden: 104 beds30), while others had limited medical facilities.50
APOD
In total, 25 studies10,13,26,29–36,38,39,41–46,49–52,54,55,57–59,61 described the APOD with 226,31 describing both APOD and PPOD. Location of APOD was gained from death certifi-cates or national registries (15),10,26,29,30,32,34,38,42–46,49,51,54,57,59 medical records or general practitioner (GP) surveys (3),31,39,61 cancer registry records (1),41 and post bereave-ment informal carer interview or survey (6).26,33,35,50,52,55,58 A total of 14 studies investigated the POD for cancer dece-dents (n = 40,451),26,29,30–34,41,45,49,51,54,57,58 6 for terminally ill or palliative decedents (n = 202,184);10,36,38,39,50,59,61 3 for all deaths (n = 30, 550; cancer or HIV/AIDS accounted for 39%–72% of deaths)35,42,43/44,55 and 2 studies reported on decedents with chronic obstructive pulmonary disease (COPD) or lung cancer.36,46 Three studies did not provide the number of rural deaths.32,57,59 A total of 11 studies were of high quality,10,13,30,34,38,41,42,45,50–52,57,58 12 medium quality26,29,32,
33,35,36,39,43,44,46,49,54,55,59 and 2 low quality.31,61 The main weaknesses were lack of rural description. Poor-quality
studies had low numbers, with one single practice study31 relying on GPs’ recollection of APOD.
Table 1 summarises the study descriptions, participant populations, study design and either percentage of rural patients who died at home (or out of hospital) or aOR of rural home death compared to urban.
The rate of home deaths ranged from 12% (Sweden30) to 81.7% (Taiwan45); 7 studies reported rates of home death less than 25% (Sweden,30 Canada,36,42,51 Australia,39,52 England10,38), 10 studies reported 25%–50% (Denmark,29 Scotland,31 England,26 Netherlands,10,38 Belgium,10,38 Australia,54 Germany,35,55 Botswana,43,44 New Zealand,61 Japan50) and 5 studies reported >50% (Mexico,32 Spain,33 Italy,34 England,58 Taiwan45). Where a home death was not achieved, most deaths occurred in hospital (New Zealand: 21.5%,61 Canada: 76.2%13,51). Less than 10% of rural deaths occurred in nursing/care homes or hospices; how-ever, these options, especially hospices, were often not mentioned29,32 or not available.13,30,33,34,39,43–45,50,54,58 Where results were expressed as aOR, compared to urban home deaths (aOR = 1), rural home deaths ranged from 0.74 (95% confidence interval (CI): 0.61–0.91; Western Australia)59 to 3.5 (95% CI: 3.03–4.04; South Australia).41
The highest rates of home deaths were seen in cancer patients (53%34 to 81%45); however, one cancer study reported the lowest frequency (12%30). Rates for termi-nally ill patients (cancer and non-cancer) ranged from 15.4%36 to 47%50 and all deaths 20.2%42 to 43%.35,55
A total of 17 studies included both urban (city or large town) and rural data. Four studies reported that rural patients are more likely to die in hospital13,51,52,57 or long-term care facility36 than urban patients with the out- of-hospital trend less marked in rural locations.51 The remaining studies reported the chance of dying at home to be greater in rural than urban locations.
Most studies designated rural residency as an inde-pendent variable with the seven rural studies briefly reporting factors influencing POD. The factors influenc-ing the decision to transfer from home included proxim-ity to a hospital or hospice,25,30,54,57,58 inability of carer to continue,54,58,61 increasing symptom burden,50,54,61 brief interval from diagnosis to death30 and patient choice.61 Two rural studies39,50 reported that the involvement of community nurses increased the likelihood of remaining home.
PPOC/PPOD
Eight studies described PPOD,26,31,37,40,48,56,60,62,63 three PPOC47,64,65 and one reported preferences for both places53 (Table 2). In all, 11 studies were prospective and 1 retrospective;31 8 quantitative,31,37,40,53,56,60,64,65 3 qualitative47,48,63 and 1 mixed.26,62 The quantitative stud-ies collected data from interviews and surveys; qualitative studies used semi-structured and open-ended interviews
at Australian National University on August 11, 2016pmj.sagepub.comDownloaded from
756 Palliative Medicine 30(8)T
able
2.
Stud
ies
desc
ribi
ng r
ural
pre
ferr
ed p
lace
of c
are
and
deat
h.
Stud
y de
scri
ptio
nR
ural
N/t
otal
NSt
udy
desi
gnPa
rtic
ipan
t/in
form
ants
Rur
al d
escr
iptio
nR
esul
tsPe
rcen
tage
(or
aO
R)
of
rura
l res
iden
ts w
hose
PP
OD
/C is
hom
e
Qua
lity
Car
roll3
1
Scot
land
1992
–199
4 an
d O
ct
1995
–Sep
tem
ber
1996
PPO
DR
ural
N =
82
(pre
-aud
it)
and
34 (
audi
t)
Ret
rosp
ectiv
e cr
oss-
sect
iona
l st
udy;
inte
rvie
ws
with
GPs
pr
e-au
dit
and
audi
t af
ter
impl
emen
tatio
n of
mat
rix
to
prom
pt G
P to
ask
PPO
D.
RR
R: p
re-a
udit
= 7
1%;
audi
t = 8
8%
Dec
eden
ts: c
ance
r pa
tient
sIn
form
ant:
GPs
Sem
i-rur
al s
mal
l to
wn,
pop
ulat
ion:
11
,240
.32
km
from
A
berd
een
PPO
D: p
re-a
udit
(aud
it): h
ome
= 7
2%
(77%
), co
mm
unity
hos
pita
l = 9
%
(21%
), N
H =
5%
(0)
, hos
pita
l = 9
% (
0),
hosp
ice
= 5
% (
3%).
Influ
enci
ng fa
ctor
s: n
ot p
rovi
ded
PPO
D: 7
2% p
re-a
udit
PPO
D: 7
7% a
udit
Med
ium
Cho
i et
al.53
Seou
l, So
uth
Kor
eaD
ates
not
giv
enPP
OC
/PPO
DR
ural
pat
ient
sN
= 7
9/37
1(21
%)
Rur
al c
arer
sN
= 4
8/28
1 (1
7%)
Cro
ss s
ectio
nal,
self-
com
plet
ed, s
ervi
ce-b
ased
su
rvey
Act
ual q
uest
ionn
aire
not
pr
ovid
ed
Can
cer
patie
nts
and
thei
r ca
rers
Not
pro
vide
dO
vera
ll PP
OC
: hom
e =
53%
pat
ient
s an
d 49
% fa
mily
Ove
rall
PPO
D: h
ome
= 4
7% p
atie
nts
and
51%
fam
ily.
Urb
an h
ome
pref
eren
ce a
OR
= 1
Influ
enci
ng fa
ctor
s: h
ome
– st
rong
er
fam
ily n
etw
ork
and
less
acc
ess
to
hosp
ital;
rura
l res
iden
cy in
depe
nden
t va
riab
le
PPO
C h
ome:
rur
al
patie
nt a
OR
= 1
.57
(95%
CI =
0.9
4–2.
61).
Rur
al fa
mily
aO
R =
1.5
7 (9
5% C
I = 0
.84–
2.94
)PP
OD
hom
e: r
ural
pa
tient
aOR
= 1
.65
(95%
C
I = 1
–2.7
2).
Rur
al fa
mily
aO
R =
2.7
6 (9
5% C
I = 1
.41–
5.42
)
Med
ium
Fore
man
et
al.56
Sout
h A
ustr
alia
, A
ustr
alia
2004
PPO
DR
ural
N =
792
/301
5 (2
6%)
Cro
ss-s
ectio
nal,
stat
e-w
ide
popu
latio
n, fa
ce-t
o-fa
ce s
urve
y (a
nnua
l hea
lth o
mbu
dsm
an
surv
ey);
stra
tifie
d by
loca
tion
of r
esid
ence
.Q
uest
ion:
whe
re w
ould
you
like
to
die
if y
ou h
ad a
ter
min
al
dise
ase,
suc
h as
can
cer
or
emph
ysem
a?
Gen
eral
po
pula
tion:
>
15 ye
ars
Rur
al: >
1000
re
side
nts
Rur
al P
POD
: hom
e =
71.
1%,
hosp
ital =
1.7
%, h
ospi
ce =
6.2
%,
NH
= 2
1%.
Urb
an: h
ome
= 6
9.9%
, hos
pita
l = 0
.4%
, ho
spic
e =
11.
4%, N
H =
18.
7%
(p <
0.0
01).
Influ
enci
ng fa
ctor
s: n
ot d
escr
ibed
; ru
ral r
esid
ency
inde
pend
ent
vari
able
PPO
D: 7
1.1%
Med
ium
Fuku
i et
al.65
Japa
nM
arch
201
0PP
OC
/PPO
DR
ural
N =
412
/104
2 (4
0%)
Cro
ss-s
ectio
nal,
natio
n-w
ide
surv
ey; s
trat
ified
by
loca
tion
of
resi
denc
e.Q
uest
ion:
pre
fere
nces
for
end-
of-li
fe c
are
and
deat
h (h
ome,
acu
te h
ospi
tal,
PCU
, pu
blic
NH
, pri
vate
NH
) if
life
expe
ctan
cy is
1–2
mon
ths,
no
phys
ical
dis
tres
s an
d ne
eded
ca
re a
ssis
tanc
e
Gen
eral
po
pula
tion:
40
–79
year
s
Rur
al a
rea:
<
150,
000
resi
dent
sPP
OC
a : ur
ban
– ho
me
= 4
9%,
hosp
ital =
17.
5%, P
CU
= 2
2.5%
, N
H =
11%
; rur
al –
hom
e =
51%
, ho
spita
l = 1
6%, P
CU
= 2
1%,
NH
= 1
2%; p
-val
ue n
ot s
igni
fican
t.O
vera
ll PP
OD
: hom
e =
43%
, ho
spita
l = 1
6%, P
CU
= 2
0%, p
ublic
N
H =
7.4
%, p
riva
te N
H =
0.6
%,
unce
rtai
n =
13.
4%.
Rur
al P
OD
dat
a no
t pr
ovid
ed.
Influ
enci
ng fa
ctor
s: n
ot d
escr
ibed
; ru
ral r
esid
ency
inde
pend
ent
vari
able
PPO
C: h
ome
= 5
1%M
ediu
m
at Australian National University on August 11, 2016pmj.sagepub.comDownloaded from
Rainsford et al. 757
Stud
y de
scri
ptio
nR
ural
N/t
otal
NSt
udy
desi
gnPa
rtic
ipan
t/in
form
ants
Rur
al d
escr
iptio
nR
esul
tsPe
rcen
tage
(or
aO
R)
of
rura
l res
iden
ts w
hose
PP
OD
/C is
hom
e
Qua
lity
Gu
et a
l.37
Shan
ghai
, Chi
naD
ecem
ber
2007
–Dec
embe
r 20
12PP
OD
Rur
al d
yads
N =
278
/522
(53
%)
Pros
pect
ive,
cro
ss-s
ectio
nal,
serv
ice-
base
d, s
epar
ate
face
-to-
face
inte
rvie
w q
uest
ionn
aire
Que
stio
n: ‘w
here
do
you
pref
er
to d
ie if
cir
cum
stan
ces
allo
wed
yo
u to
cho
ose
at y
our
term
inal
st
age
of li
fe?’
PPO
D: h
ome
or h
ospi
tal
Can
cer
patie
nt/
care
r dy
adN
ot p
rovi
ded
Rur
al p
atie
nt P
POD
a : ho
me
60.4
%
(urb
an 4
5.9%
), ho
spita
l 32%
, fam
ily t
o de
cide
7%
.C
arer
: ove
rall
pref
eren
ce fo
r ho
me
= 5
1.3%
(no
rur
al %
pro
vide
d).
Influ
enci
ng fa
ctor
s: c
ultu
re; r
ural
re
side
ncy
inde
pend
ent
vari
able
PPO
D: 6
0.4%
Med
ium
How
ell e
t al
.40C
anad
aIm
pact
of s
hare
d ca
re
mod
el o
n PO
DR
ural
N =
95
Pros
pect
ive
coho
rt s
tudy
; fa
ce-t
o-fa
ce in
terv
iew
qu
estio
nnai
re; p
refe
renc
es a
t tim
e of
enr
olm
ent
into
stu
dy.
RR
: 83%
Que
stio
n: ‘i
f you
r co
nditi
on
wor
sens
, wha
t is
you
r ho
pe
abou
t w
here
you
wou
ld li
ke
to r
ecei
ve c
are?
+ P
OD
(if
appr
opri
ate)
’
Patie
nts;
adv
ance
d,
prog
ress
ive
dise
ase
(mal
igna
nt
and
non-
mal
igna
nt)
Rur
al m
unic
ipal
ity
com
pris
ing
smal
l tow
ns,
popu
latio
n =
80,
000,
se
rved
by
60-b
ed
com
mun
ity h
ospi
tal
PPO
D: h
ome
= 8
0%, h
ospi
tal =
4%
, ho
spic
e =
1%
PPO
C d
ata
not
repo
rted
Influ
enci
ng fa
ctor
s: n
ot p
rovi
ded
as
POD
was
sec
onda
ry o
utco
me
PPO
D: 8
0%H
igh
McC
all a
nd R
ice4
7
Wes
t H
ighl
ands
, Sc
otla
ndPO
C d
ecis
ions
Rur
al N
= 8
Pros
pect
ive,
qua
litat
ive,
pu
rpos
ive
sam
ple
iden
tifie
d by
GP;
sem
i-str
uctu
red-
tape
d in
terv
iew
s.Q
uest
ion:
‘if e
very
thin
g w
ere
poss
ible
whe
re w
ould
you
ch
oose
to
be c
ared
for
in t
he
end
stag
e of
you
r ill
ness
?’ an
d ’w
hy w
ould
you
cho
ose
this
pl
ace?
’
Can
cer
patie
nts
Not
pro
vide
dR
ural
PPO
C: h
ome
= 2;
no
pref
eren
ce
but
hom
e no
t po
ssib
le =
2; h
ospi
tal a
s ho
me
not
poss
ible
= 4
; PPO
C c
hang
es
as c
are
need
s in
crea
se.
Influ
enci
ng fa
ctor
s: c
arer
ava
ilabi
lity
and
abili
ty p
atie
nts’
wis
hes
may
not
be
the
ir p
refe
renc
e du
e to
con
cern
ab
out
care
r an
d no
t w
antin
g to
be
com
e a
burd
en.
Det
erm
inan
ts fo
r PP
OC
: ca
rer
reso
urce
and
sup
port
; pa
st e
xper
ienc
es w
ith d
eath
; co
mm
unic
atio
n of
wis
hes
PPO
C: 2
5%H
igh
McG
rath
48
Nor
ther
n T
erri
tory
of
Aus
tral
ia.
PPO
DN
= 7
2
Qua
litat
ive,
phe
nom
enol
ogic
al
stud
y, o
pen-
ende
d in
terv
iew
s w
ith a
cro
ss-s
ectio
n of
pa
rtic
ipan
ts.
Que
stio
n no
t pr
ovid
ed
Indi
geno
us
patie
nts
(n =
10)
, ca
regi
vers
(n
= 1
9);
Indi
geno
us a
nd
non-
Indi
geno
us
heal
th c
are
wor
kers
(n
= 4
1),
inte
rpre
ters
(n =
2)
Rur
al a
nd r
emot
e te
rrito
ry o
f Aus
tral
iaC
lear
art
icul
atio
n of
Indi
geno
us
peop
le t
o di
e at
hom
e co
nnec
ted
to
land
and
fam
ily.
Influ
enci
ng fa
ctor
s: s
tron
g cu
ltura
l re
ason
s gi
ven
incl
udin
g st
rong
co
nnec
tion
with
land
and
com
mun
ity,
a be
lief i
n ‘d
eath
cou
ntry
’
Hom
eM
ediu
m
(Con
tinue
d)
Tab
le 2
. (C
ontin
ued)
at Australian National University on August 11, 2016pmj.sagepub.comDownloaded from
758 Palliative Medicine 30(8)
Stud
y de
scri
ptio
nR
ural
N/t
otal
NSt
udy
desi
gnPa
rtic
ipan
t/in
form
ants
Rur
al d
escr
iptio
nR
esul
tsPe
rcen
tage
(or
aO
R)
of
rura
l res
iden
ts w
hose
PP
OD
/C is
hom
e
Qua
lity
Shih
et
al.60
Tai
wan
Apr
il 20
11PP
OC
/PPO
DR
ural
pat
ient
s N
= 7
6/31
0 (2
5%)
Cro
ss-s
ectio
nal;
natio
n-w
ide
post
al q
uest
ionn
aire
; str
atifi
ed
by lo
catio
n of
res
iden
ce.
Asse
ssm
ent:
five-
part
qu
estio
nnai
re in
clud
ed P
POC
/PP
OD
(ho
me
or n
ot h
ome)
bo
th in
cur
rent
situ
atio
n an
d hy
poth
etic
al if
phy
sici
ans
or
hom
e nu
rsin
g ca
re s
ervi
ces
wer
e av
aila
ble
Reg
iste
red
gene
ral
patie
nts
Not
pro
vide
d.
Res
iden
tial
cate
gori
es w
ere
city
, to
wn,
rur
al
Rur
al p
atie
nts
PPO
Da :
hom
e =
79%
, no
t ho
me
= 2
1%C
ity p
atie
nt P
POD
: hom
e =
57%
, not
ho
me
= 4
3%; p
= 0
.005
.R
ural
PPO
C n
ot r
epor
ted.
Influ
enci
ng fa
ctor
s: n
ot d
escr
ibed
; ru
ral r
esid
ency
inde
pend
ent
vari
able
PPO
D: 7
9%M
ediu
m
Tho
mas
et
al.26
,62
Nor
th W
est
Engl
and,
U
KO
ctob
er
2000
–Sep
tem
ber
2002
PPO
DR
ural
pat
ient
s: v
illag
e/ha
mle
t N
= 1
2; s
mal
l to
wn
N =
6/4
1R
ural
car
er: (
Ken
dall)
N
= 6
/18
2-ye
ar lo
ngitu
dina
l ob
serv
atio
nal s
tudy
; fac
e-to
-fa
ce in
terv
iew
s.Pr
ompt
que
stio
n: ‘i
f you
r ill
ness
ge
ts w
orse
, whe
re w
ould
you
lik
e to
be?
’ Wou
ld y
ou li
ke, f
or
exam
ple,
to
go t
o a
hosp
ital,
be
at h
ome,
go
to a
frie
nd/fa
mily
, go
to
hosp
ice,
go
to a
NH
, sta
y w
here
you
are
now
/’ (s
uppl
ied
in fu
ll re
port
)
Can
cer
patie
nts
(>18
year
s;
3 m
onth
s lif
e ex
pect
ancy
)
Sem
i-rur
al S
outh
La
kela
nd (
mar
ket
tow
n K
enda
ll-sm
all
hosp
ital);
villa
ge/h
amle
t an
d sm
all t
own
PPO
D: v
illag
e/ha
mle
t: ho
me
= 2
5%,
hosp
ice
= 1
7%, h
ospi
tal =
0,
NH
= 8
%, h
ome
or h
ospi
ce =
33%
, un
deci
ded
= 1
7%.
Smal
l vill
age:
hom
e =
17%
, ho
spic
e =
50%
, und
ecid
ed =
33%
.In
fluen
cing
fact
ors:
ove
rall
fact
ors
pres
ente
d bu
t no
t ru
ral s
peci
fic; r
ural
re
side
ncy
inde
pend
ent
vari
able
PPO
D: h
amle
t = 2
5%;
villa
ge =
17%
Hig
h
Vei
llett
e et
al.6
3
Que
bec,
Can
ada
Janu
ary–
Apr
il 20
07‘G
ood
deat
h’R
ural
N =
40
Qua
litat
ive,
eth
nogr
aphi
c st
udy
conc
eptu
alis
e th
e go
od r
ural
de
ath;
indi
vidu
al in
terv
iew
s an
d fo
cus
grou
ps.
End-
of-li
fe lo
catio
ns: h
ome
and
hosp
ital
Com
mun
ity
mem
bers
with
ex
peri
ence
of
deat
h an
d dy
ing
in
rura
l com
mun
ities
Reg
ion
A: m
etro
-ad
jace
nt t
o Q
uebe
c ci
ty.
Reg
ion
B:
cons
ider
able
di
stan
ce fr
om
Que
bec
city
PPO
D: h
ome
com
mun
ity s
urro
unde
d by
fam
ily, a
nd fr
iend
s; s
olid
arity
and
su
ppor
t of
nei
ghbo
urs,
but
aw
are
of
need
for
avai
labl
e he
alth
car
e an
d su
ppor
t se
rvic
es.
Seve
ral w
ould
pre
fer
less
car
e an
d fe
wer
ser
vice
s th
an fo
rgo
cont
act
with
fam
ily a
nd fr
iend
s
Hom
e co
mm
unity
Hig
h
Wils
on e
t al
.64A
lber
ta, C
anad
aM
ay–J
uly
2010
PPO
CR
ural
N =
213
/120
3 (1
7.7%
)
Ann
ual p
rovi
nce,
wid
e po
pula
tion–
base
d te
leph
one
surv
ey; s
trat
ified
by
loca
tion
of
resi
denc
e.Q
uest
ion:
‘whe
re w
ould
you
ch
oose
to
spen
d yo
ur la
st d
ays
of li
fe?
Gen
eral
pub
lic:
>18
year
sN
ot p
rovi
ded
PPO
C: r
ural
– h
ome
= 7
4.9%
, ho
spic
e =
11.
5%, N
H =
1.1
%,
hosp
ital =
8.2
%, u
nkno
wn
= 4
.4%
Vill
age:
hom
e =
53.
3%, h
ospi
ce =
20%
, N
H =
3.3
%, h
ospi
tal =
16.
7%,
unkn
own
= 6
.7%
.C
ity: h
ome
= 7
0.3%
, hos
pice
= 1
5.5%
, N
H =
1.8
%, h
ospi
tal =
6.1
%; p
= 0
.4.
Influ
enci
ng fa
ctor
s: n
ot r
ural
spe
cific
; ru
ral r
esid
ency
inde
pend
ent
vari
able
PPO
C: r
ural
: 74.
9%;
villa
ge: 5
3.3%
Med
ium
POD
: pla
ce o
f dea
th; P
OC
: pla
ce o
f car
e; P
POC
/PPO
D: p
refe
rred
pla
ce o
f car
e/de
ath;
NH
: nur
sing
hom
e; P
CU
: pal
liativ
e ca
re u
nit;
RR
R: r
ural
res
pons
e ra
te; R
R: o
vera
ll re
spon
se r
ate;
aO
R: a
djus
ted
odds
rat
io.
a Dat
a re
calc
ulat
ed b
y re
sear
cher
to
enab
le r
ural
PO
D p
erce
ntag
es t
o be
rep
orte
d.
Tab
le 2
. (C
ontin
ued)
at Australian National University on August 11, 2016pmj.sagepub.comDownloaded from
Rainsford et al. 759
and focus groups. Preferences were collected for 2531 rural residents, the smallest study47 interviewed eight cancer patients and the largest56 surveyed 792 members of the general public. Preferences were collected from 680 patients (73% cancer), 351 family carers, 43 health professionals and 1457 members of the general public. Studies were conducted in eight countries (Canada,40,63,64 Australia,48,56 Scotland,31,47 China,37 South Korea,53 Taiwan,60 Japan65 and England26,62). One setting was described as rural and remote,48 three rural,40,47,63 one semi-rural31 and seven population-based studies included rural participants.26,37,53,56,60,62,64,65 Most studies did not pro-vide details of the rural setting only stating rural partici-pants resided in small towns, villages and hamlets or rural areas either adjacent to or some distance from a major city. One qualitative study (phenomenological) explored PPOD for rural and remote Aboriginal people in the Northern Territory of Australia,48 another (ethnographic) aimed to conceptualise the good death for Quebec rural franco-phones.63 One UK study explored factors influencing deci-sions around POC in the West Highlands of Scotland,47 and the other examined the reasons behind cancer patients’ POD preferences in North West England.62
Four studies (one quantitative40 and three qualita-tive26,47,62,63) were of high quality, and the remainder stud-ies were of medium quality. The main weaknesses were inadequate rural descriptions, omission of rural response rate and uncertainty as to generalisability of results to other rural locations as participants were either recruited from single health services or members of a specific cul-ture (remote Indigenous Australians). All qualitative stud-ies had clear aims and justifications for study design and reflected on researcher biases.
Eight studies provided the actual question asked to assess preferences.26,37,40,47,56,60,62,64,65 One questionnaire offered home/not home as the option,60 another home or hospital37 with the remainder being either open-ended or providing a list of options. Cancer patients were asked their preferences if there were no influencing factors such as physical symptoms, carer burden or unlimited provision of home support.
Preference for dying at home
Six quantitative studies reported the preference for dying at home as a percentage,26,31,37,40,56,60,62 and one study reported as an aOR compared to urban preferences.53 One qualitative study reported both percentages and themes.47 All studies, with the exception of two UK studies,47,62 reported more than 50% of rural participants expressed a preference for dying at home. Thomas et al.26 reported home to be the preferred place in 25% of participants; however, an additional 33% of participants selected ‘home or hospice’ as their preferred place, an option not available in other studies. Home was the PPOD for 60.4%37 to 80%40
of patients and 71.1%56 to 79%60 of the general public. No statistical results were reported for carers. In one study, home was the PPOC for 25%47 of patients with home as the PPOC for 51%65 to 74.9%64 of the general public.
Three studies53,64,65 reported both urban and rural pref-erences for POC as either a percentage or odds ratio with a fourth study60 reporting PPOD as a percentage. Results reported that preferences for home were greater in rural than urban areas and only the preferences for POD were statistically significant.
There was clear articulation in all qualitative studies that home was the PPOD ‘surrounded by family, and friends’63 and in the case of Indigenous Australians ‘connected to land and family’.48 The qualitative studies provided mean-ing behind decision-making, acknowledging that decision-making is complex and ‘wishes’ and ‘preferences’ are not necessarily the same. While wishing to remain at home, the participants in McCall and Rice’s study47 were aware that achieving this may not be possible due to ‘concern about their carer and not wanting to become a burden’, and so hospital became their preference.
Preference for hospital
With the exception of one study that recorded options as ‘home’ or ‘not home’,60 all participants were given the option for hospital as an alternate POD or POC. The pref-erence for hospital ranged from less than 4% (England,26,62 Australia,56 Canada40) to 32% (China37). Reasons for pre-ferring hospital included perceptions of better professional care, relief of carer burden and symptom control. Familiarity with hospitals was also reported. Hospitals were not always available within the community, and dis-placement from community to access the closest hospital was not always acceptable with some participants prefer-ring to forgo symptom management to remain in their community63.
Preference for hospice
Six studies25,37,47,48,56,60 acknowledged that inpatient hos-pices were not available in all locations, therefore requir-ing patients to travel outside their immediate community or were not offered as an option. Five studies reported preferences for inpatient hospice care with preferences ranging from 1% (Canada40), 17% (North West England26,62) to 21% in Japan.65
Preference for residential aged care facility
Only 5 studies26,31,56,62,64,65 included residential aged care facilities (RACFs or Nursing Homes) as an option with preferences ranging from 1.1% in Canada64 to 21% in South Australia.56 Many rural locations did not have aged care facilities.
at Australian National University on August 11, 2016pmj.sagepub.comDownloaded from
760 Palliative Medicine 30(8)
Discussion
To the best of our knowledge, this is the first systematic review of the international literature describing the actual and PPOC/PPOD of rural residents with most rural data embedded in population reports rather than in specific rural studies. The review also highlights the variability in the use of the term ‘rural’. Due to heterogeneity in rural locations, individual studies could be considered ‘silos’; however, consistency of results was present across a num-ber of studies with death certificates documenting hospital to be the most frequent POD. Results described home to be the PPOC/PPOD; however, the small qualitative stud-ies26,47,62,63 highlight the complexity and fluidity surround-ing preferences indicating it is not simply a hypothetical decision as suggested by population surveys.
The APOD studies provided a ‘snap shot’ of current practice by describing and comparing POD across nations through death certificate data, while acknowledging that many developing nations are not represented. While the cause of death (cancer/non-cancer) stated on the certificate could be different to that found at autopsy,9,51 any miss-reporting of POD is unlikely to be significant as many studies included a cross-checking process. In these large studies, rural regions were included as either part of the whole country/state/province or stratified by degree of urbanisation; however, rural response rates were mostly unreported but inferred to be at or close to 100%. Common to previous studies, the rural data do not consider POC, and while hospital is the documented POD, decedents may have spent a significant time at home only to be transferred to hospital at the very end of life.6
Consistent with urban studies,8,66 most rural residents do have a preference for POC/POD, often home; however, high quality studies reported both the highest and lowest preference rates for a home death. For some, due to remote-ness, there may not be a choice as there is no local alterna-tive to home50 or the community hospital is underserviced.37 Rather than a desire to remain home for whatever reason it may be more an ‘aversion to hospital’.47 Thomas et al.62 and McCall and Rice47 observed that there is a difference between ‘wish’ and ‘preference’ with preferences chang-ing to accomplish end-of-life care goals67 as death approaches, care needs increase or carer burden becomes apparent.68 Some patients are willing to sacrifice symptom management to remain home, while others sacrifice their preferred place to reduce carer burden. While surveys pro-vide hypothetical preferences, there is an apparent signifi-cant incongruence between actual and preferred POC/POD consistent with urban data. However, reasons behind and absolute importance of preferences are not adequately described, and care must be taken when reporting the pro-portion of rural people who do not die in their preferred place as one cannot compare the hypothetical preferences of healthy population survey participants with terminally ill patients and their carers who are living the experience.
Factors influencing POC/POD in rural settings were not clearly identified as ‘place’ was the dependent variable and rural residency one of many independent variables. When reported, influencing factors were consistent with urban findings and included functional status and clinical condi-tion, carer and social networks and health-system facili-ties; however, one needs to be careful drawing conclusions due to the limited rural data. Rural resilience52 and com-munity as a social support network69,70 were not described. The uniqueness of individual rural communities was evi-dent with locations exhibiting the highest numbers of or strongest preferences for home deaths being those with stronger traditional cultures and values; however, these locations often have less access to hospitals, hospices and nursing homes.
Previous studies report POC and POD are not synony-mous,71 POD is not necessarily a dominant concern of the dying person72 and a hospital death is not necessarily a ‘bad death’68 with one study73 suggesting rural residents are disadvantaged by the lack of inpatient facilities. Hospices can be an acceptable alternative to home;8,26,62 however, no two rural locations are the same and not all rural residents have access73 to an inpatient hospice. When available, the rural community hospital often becomes the alternative to hospice, albeit without specialist multidisciplinary teams, and may be one reason for the higher rate of rural palliative care deaths in hospital in some locations.
If each rural community is unique, the challenge for future research is to find a common thread while avoiding over-generalisation.73 However, trying to respect the wishes of each patient and community, it is not always practical or financially viable to have multiple individual models to suit each location. While ensuring all rural communities are represented using a stratified method of sampling, it is also important to ensure a cross-section of voices are heard within each represented rural community. Future studies could explore the actual proportion of rural residents who die in their preferred place and include qualitative studies with people (patients, carers, community members, rural health professionals) experienced in end-of-life care. While it is important to increase community services to support those who wish to stay at home, improvements in palliative care within community hospitals74,75 and RACF should not be overlooked as these places are often substitute hospice and the only alternative to home.
Study limitations
This systematic review had limitations in both complete-ness of the review and quality of identified studies. Studies not including selection criteria wording in the title, abstract or keywords may have been missed; however, electronic searching was augmented by hand-searching journals and reference lists. The lack of homogeneity and consistency in rural definition meant that studies reported on a variety of rural locations, populations and proximity to health
at Australian National University on August 11, 2016pmj.sagepub.comDownloaded from
Rainsford et al. 761
services with significant differences found in POD not only between regions but within regions.34,57 The objec-tives of most studies were not rural focused, and in some studies, POD was a secondary outcome increasing the risk of misinterpretation or miscalculation of rural data. Other limitations include small rural sample size; absence of documented rural response rate; limited generalisability due to specific populations such as remote indigenous peo-ples; recruitment bias (single medical practices or hospital clinics, omission of rural patients unable to travel to city cancer clinics and those not associated with palliative care services); reliance on family and health professionals’ recall of patient preferences; variations in terminology and options for POD; inconsistency of nursing homes as ‘home’ or ‘institution’; inconsistency in cohorts (cancer/non-cancer/all deaths/general population).
Conclusion
Despite striking variations between and within countries, a significant number of rural residents die in hospital; how-ever, in many locations, they have a greater chance of dying at home than city residents. The PPOD (and care) for most rural residents is home; however, home is not the first preference for all and more work is needed to explore influencing factors, changes in preferences as disease pro-gresses, the absolute importance of preferences, the influ-ence of community on POD and the experience of providing and receiving palliative care in rural hospitals.
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
References
1. Patrick D, Engelberg R and Randall Curtis J. Evaluating the quality of dying and death. J Pain Symptom Manage 2001; 22(3): 717–726.
2. Hunt KJ, Shlomo N and Addington-Hall J. End-of-life care and achieving preferences for place of death in England: results of a population-based survey using the VOICES-SF questionnaire. Palliat Med 2014; 28(5): 412–421.
3. Higginson I and Sen-Gupta G. Place of care in advanced cancer: a qualitative systematic literature review of patient preferences. J Palliat Med 2000; 3(3): 287–300.
4. Kellehear A. A social history of dying. Cambridge: Cambridge University Press, 2008.
5. Singer M, Martin D and Kelner M. Quality end-of-life care patients’ perspectives. JAMA 1999; 281(2): 163–168.
6. Gomes B, Calanzani N, Gysels M, et al. Heterogeneity and changes in preferences for dying at home: a systematic review. BMC Palliat Care 2013; 12: 7.
7. Cohen J, Bilsen J, Addington-Hall J, et al. Population-based study of dying in hospital in six European countries. Palliat Med 2008; 22: 702–710.
8. Gomes B, Calanzani N and Higginson I. Local preferences and place of death in regions within England 2010. Report, August 2011. London: Cicely Saunders International, http://www.kcl.ac.uk/lsm/research/divisions/cicelysaunders/attach-ments/keyreport-Local-preferences-and-place-of-death-in-regions-within-England.pdf (accessed 11 May 2015).
9. Cohen J, Bilsen J, Addington-Hall J, et al. Population based study of dying in hospital in six European countries. Palliat Med 2008; 22: 702–710.
10. Cohen J, Houttekier D, Onwuteaka B, et al. Which patients with cancer die at home? A study of six European coun-tries using death certificate data. J Clin Onc 2010; 28(13): 2267–2273.
11. Broad J, Gott M, Hongsoo K, et al. Where do people die? An international comparison of the percentage of deaths occurring in hospital and residential aged care settings in 45 populations, using published and available statistics. Int J Public Health 2013; 58: 257–267.
12. Wilson DM, Truman CD, Thomas R, et al. The rapidly changing location of death in Canada, 1994–2004. Soc Sci Med 2009; 68(10): 1752–1758.
13. Burge F, Lawson B and Johnston G. Trends in the place of death of cancer patients, 1992–1997. CMAJ 2003; 168(3): 265–270.
14. Houttekier D, Cohen J, Surkyn J, et al. Study of recent and future trends in place of death in Belgium using death cer-tificate data. BMC Public Health 2011; 11: 228.
15. Mystakidou K, Parpa E, Tsilika E, et al. Where do cancer patients die in Greece? A population-based study on the place of death in 1993 and 2003. J Pain Symptom Manage 2009; 38(2): 309–314.
16. Yamagishi A, Morita T and Miyashita M. Preferred place of care and place of death of the general public and cancer patients in Japan. Support Care Cancer 2012; 20: 2575–2582.
17. Tang S, Wu S, Hung Y, et al. Trends in quality of end-of-life care for Taiwanese cancer patients who died in 2000–2006. Ann Oncol 2009; 20: 343–348.
18. Gomes B and Higginson I. Where people die (1974–2030): past trends, future projections and implications for care. Palliat Med 2008; 22: 33–41.
19. Australian Bureau of Statistics. Year book Australia, 2012, http://www.abs.gov.au/ausstats/abs@.nsf/Lookup/by%20Subject/1301.0~2012~Main%20Features~Geographic%20distribution%20of%20the%20population~49 (accessed 13 June 2015).
20. Australian Institute of Health and Welfare. Rural, regional and remote health: a guide to remoteness classifications. AIHW catalogue no. PHE 53, 2004. Canberra, ACT, Australia: AIHW, http://www.aihw.gov.au/WorkArea/DownloadAsset.aspx?id=6442459567 (accessed 11 October 2015).
21. Spice R, Paul L and Biondo P. Development of a rural pal-liative care program in the Calgary zone of Alberta Health services. J Pain Symptom Manage 2012; 43(5): 911–924.
22. Wilson DM, Fillion L, Thomas R, et al. The ‘good’ rural death: a report of an ethnographic study in Alberta, Canada. J Palliat Care 2009; 25(1): 21–29.
at Australian National University on August 11, 2016pmj.sagepub.comDownloaded from
762 Palliative Medicine 30(8)
23. Commonwealth of Australia. Supporting Australians to live well at the end of life: National Palliative Care Strategy 2010. Canberra, ACT, Australia: Commonwealth of Australia, 2010.
24. NSW Ministry of Health. The NSW Government plan to increase access to palliative care 2012–2016. Sydney, NSW, Australia: NSW Ministry of Health, 2012.
25. Moher D, Liberati A, Tetzlaff J, et al. Preferred report-ing items for systematic reviews and meta-analyses: the PRISMA statement. Ann Intern Med 2009; 151(4): 264–270.
26. Thomas C, Morris S and Gatrell T. Place of death in the Morecambe Bay area: patterns of and preferences for place of final care and death among terminally ill cancer patients and their carer. Final report to The Research and Development Department, NHS Directorate of Health and Social Care North. Reference RDO/28/3/10, January 2003. Lancaster: University of Lancaster.
27. Hawker S, Payne S, Kerr C, et al. Appraising the evidence: reviewing disparate data systematically. Qual Health Res 2002; 12(9): 1284–1299.
28. Kitto S, Chesters J and Grbich C. Quality in qualitative research. Criteria for authors and assessors in the submis-sion and assessment of qualitative research articles for the Medical Journal of Australia. Med J Australia 2008; 188(4): 243–246.
29. Aabom B, Kragstrup J, Vondeling H, et al. Population-based study of place of death of patients with cancer: implications for GPs. Br J Gen Pract 2005; 55(518): 684–689.
30. Axelsson B and Christensen SB. Place of death correlated to sociodemographic factors. A study of 203 patients dying of cancer in a rural Swedish county in 1990. Palliat Med 1996; 10(4): 329–335.
31. Carroll D. An audit of place of death of cancer patients in a semi-rural Scottish practice. Palliat Med 1998; 12(1): 51–53.
32. Castillo-Guzman S, Palacios-Rios D, Nava-Obregon TA, et al. Home versus hospital mortality from cancer in Mexico (1999–2009). Am J Hosp Palliat Care 2013; 30(3): 249–252.
33. Catalan-Fernandez JG, Pons-Sureda O, Recober-Martinez A, et al. Dying of cancer. The place of death and family circumstances. Med Care 1991; 29(9): 841–852.
34. Costantini M, Balzi D, Garronec E, et al. Geographical vari-ations of place of death among Italian communities suggest an inappropriate hospital use in the terminal phase of cancer disease. Public Health 2000; 114(1): 15–20.
35. Escobar Pinzon LC, Claus M, Zepf KI, et al. Preference for place of death in Germany. J Palliat Med 2011; 14(10): 1097–1103.
36. Goodridge D, Lawson J, Rennie D, et al. Rural/urban dif-ferences in health care utilization and place of death for per-sons with respiratory illness in the last year of life. Rural Remote Health 2010; 10(2): 1349, http://www.rrh.org.au
37. Gu X, Cheng W, Cheng M, et al. The preference of place of death and its predictors among terminally ill patients with cancer and their caregivers in China. Am J Hosp Palliat Care. Epub ahead of print 9 July 2014. DOI: 10.1177/1049909114542647.
38. Houttekier D, Cohen J, Bilsen J, et al. Place of death in metropolitan regions: metropolitan versus non-metropolitan
variation in place of death in Belgium, The Netherlands and England. Health Place 2010; 16(1): 132–139.
39. Howat A, Veitch C and Cairns W. A retrospective review of place of death of palliative care patients in regional north Queensland. Palliat Med 2007; 21(1): 41–47.
40. Howell D, Marshall D, Brazil K, et al. A shared care model pilot for palliative home care in a rural area: impact on symptoms, distress, and place of death. J Pain Symptom Manage 2011; 42(1): 60–75.
41. Hunt RW, Fazekas BS, Luke CG, et al. Where patients with cancer die in South Australia, 1990–1999: a population-based review. Med J Aust 2001; 175(10): 526–529.
42. Jayaraman J and Joseph K. Determinants of place of death: a population-based retrospective cohort study. BMC Palliat Care 2013; 12: 19.
43. Lazenby J and Olshevski J. Place of death among Botswana’s oldest old. Omega J Death Dying 2012; 65(3): 173–187.
44. Lazenby M, Ma T, Moffat HJ, et al. Influences on place of death in Botswana. Palliat Support Care 2010; 8(2): 177–185.
45. Lin HC and Lin CC. A population based study on the spe-cific locations of cancer deaths in Taiwan, 1997–2003. Support Care Cancer 2007; 15(12): 1333–1339.
46. Lopez-Campos JL, Ruiz-Ramos M, et al. Characteristics of subjects who died of chronic obstructive pulmonary disease in Andalusia in 2010 and 2011. J Palliat Med 2013; 16(12): 1610–1613.
47. McCall K and Rice AM. What influences decisions around the place of care for terminally ill cancer patients? Int J Palliat Nurs 2005; 11(10): 541–547.
48. McGrath P. ‘I don’t want to be in that big city; this is my country here’: research findings on Aboriginal peoples’ preference to die at home. Aust J Rural Health 2007; 15(4): 264–268.
49. Papke J and Koch R. Places of death from cancer in a rural location. Onkologie 2007; 30(3): 105–108.
50. Yasumura S, Haga H, Shibata H, et al. Factors relating to place of death of Japanese people from a small town in a rural area. Aging Clin Exp Res 2000; 12(6): 449–454.
51. Burge F. Where a cancer patient dies: effect of rural resi-dency. J Rural Health 2005; 21(3): 233–238.
52. Burns C, Dal Grande E, Tieman J, et al. Who provides care for people dying of cancer? A comparison of a rural and metropolitan cohort of South Australian bereaved popula-tion study. Aust J Rural Health 2015; 23: 24–31.
53. Choi KS, Chae YM, Lee CG, et al. Factors influencing preferences for place of terminal care and of death among cancer patients and their families in Korea. Support Care Cancer 2005; 13(8): 565–572.
54. Crawford G. Palliative care on Kangaroo Island. Aust J Rural Health 2000; 8: 35–40.
55. Escobar Pinzon LC, Weber M, Claus M, et al. Factors influ-encing place of death in Germany. J Pain Symptom Manage 2011; 41(5): 893–903.
56. Foreman L, Hunt R, Luke C, et al. Factors predictive of preferred place of death in the general population of South Australia. Palliat Med 2006; 20: 447–453.
57. Gatrell AC HJ, Francis BJ, Thomas C, et al. Place of death: analysis of cancer deaths in part of North West England. J Public Health Med 2003; 25(1): 53–58.
at Australian National University on August 11, 2016pmj.sagepub.comDownloaded from
Rainsford et al. 763
58. Herd E. Terminal Care in a semi-rural area. Br J Gen Pract 1990; 40: 248–251.
59. McNamara B. Factors affecting place of death in Western Australia. Health Place 2007; 13: 356–367.
60. Shih C, Hu WY, Cheng S, et al. patient preferences ver-sus family physicians’ perceptions regarding the place of end-of-life care and death: a nationwide study in Taiwan. J Palliat Med 2015; 18(7): 625–630.
61. Smyth D, Farnell A, Dutu G, et al. Palliative care provision by rural general practitioners in New Zealand. J Palliat Med 2010; 13(3): 247–250.
62. Thomas C, Morris SM and Clark D. Place of death: prefer-ences among cancer patients and their carers. Soc Sci Med 2004; 58: 2431–2444.
63. Viellette A, Fillion L, Wilson D, et al. La belle mort en milieu rural: a report of an ethnographic study of the good death for Quebec rural francophones. J Palliat Care 2010; 26(3): 159–166.
64. Wilson DM CJ, Deliens L, et al. The preferred place of last days: results of a representative population-based public survey. J Palliat Med 2013; 16(5): 502–508.
65. Fukui S, Yoshiuchi K, Fujita J, et al. Japanese people’s pref-erence for place of end-of-life care and death: a population based nationwide survey. J Pain Symptom Manage 2011; 42(6): 882–892.
66. Sandsdalen T, Hov R, Hoye S, et al. Patients’ preferences in palliative care: a systematic mixed studies review. Palliat Med 2015; 29(5): 399–419.
67. Evans W, Cutson T, et al. Is there no place like home? Caregivers recall reasons for and experience upon transfer
from home hospice to inpatient facilities. J Palliat Med 2006; 9(1): 100–110.
68. Gott M, Seymour J, Bellamy G, et al. Older people’s views about home as a place of care at the end of life. Palliat Med 2004; 18: 460–467.
69. Leonard R, Horsfall D and Rosenberg J. The positioning of family, friends, community, and service providers in sup-port networks for caring at end-of-life: a social network analysis. Support Palliat Care 2015; 5(Suppl. 1): A6–A7.
70. Horsfall D, Noonan K and Leonard R. Bringing our dying home: how caring for someone at the end of life builds social capital and develops compassionate communities. Health Sociol Review 2012; 21(4): 373–382.
71. Agar M, Currow D and Shelby-James TM. Preference for place of care and place of death in palliative care: are these different questions? Palliat Med 2008; 22: 787–795.
72. Steinhauser K, Christakis N, Clipp E, et al. Factors consid-ered important at the end of life by patients, family, phy-sicians and other care providers. JAMA 2000; 284(19): 2476–2482.
73. Menec V, Nowicki S and Kalischuk A. Transfers to acute care hospital at the end of life: do rural/remote regions dif-fer from urban regions. Rural Rem Health 2010; 10: 1281, http://www.rrh.org.au
74. Payne S, Kerr C, Hawker S, et al. Community hospitals: an under-recognised resource for palliative care. J R Soc Med 2004; 97: 428–431.
75. Fink R, Oman K, Youngworth J, et al. A palliative care needs assessment of rural hospitals. J Palliat Med 2013; 16(6): 638–644.
at Australian National University on August 11, 2016pmj.sagepub.comDownloaded from
Review
Rural residents’ perspectives on the rural ‘good death’: a scoping review
Suzanne Rainsford MBBS FAChPM1, Roderick D. MacLeod PhD FAChPM
2,3, Nicholas J. Glasgow MD FAChPM1,
Donna M. Wilson RN PhD4, Christine B. Phillips MPH FRACGP
1 and Robert B. Wiles MBBS FRCGP5
1Medical School, Australian National University, Canberra, Australian Capital Territory, Australia, 2HammondCare,
Sydney, New South Wales, Australia, 3Palliative Medicine, University of Sydney, Sydney, New South Wales, Australia,4Faculty of Nursing, University of Alberta, Edmonton, Alberta, Canada and 5Rural Clinical School, Australian National
University, Cooma, New South Wales, Australia
Accepted for publication 10 August 2016
CorrespondenceSuzanne RainsfordC/- PO Box 159Cooma, New South Wales 2630AustraliaE-mail: suzanne.rainsford@anu.edu.au
What is known about this topic
• The ‘good death’ is a key objectiveof palliative care.
• Current viewpoints and researchabout ‘good’ deaths reflect theurban voice.
• Rural areas are distinct and assuch need to have specialconsideration.
What this paper adds
• Rural residents in both developedand developing countries reportsimilar themes for a ‘good death’as urbanites.
• Rural perspectives are subjectiveand dependent on rural context.
• Highlights the need for moreresearch regarding rural patientand family caregiver perspectiveson what it means to die well.
AbstractThe ‘good death’ is one objective of palliative care, with many ‘gooddeath’ viewpoints and research findings reflecting the urban voice. Ruralareas are distinct and need special consideration. This scoping reviewidentified and charted current research knowledge on the ‘good’ ruraldeath through the perspectives of rural residents, including rural patientswith a life-limiting illness, to identify evidence and gaps in the literaturefor future studies. A comprehensive literature search of English languagearticles (no date filter applied) was conducted in 2016 (2 January to 14February) using five library databases. Reference lists of included articles,recent issues of eight relevant journals and three grey literature databaseswere also hand-searched. Twenty articles (for 17 studies and onesystematic review) were identified after a two-phase screening process bytwo reviewers, using pre-determined inclusion criteria. Data from eachstudy were extracted and charted, analysed using a thematic analysis ofthe included articles’ content, and with a quantitative analysis of thescoping review. These papers revealed data collected from rural patientswith a life-limiting illness and family caregivers, rural healthcareproviders, the wider rural community, rural community leaders and ruralhealth administrators and policy makers. Rural locations wereheterogeneous. Residents from developed and developing countriesbelieve a ‘good death’ is one that is peaceful, free of pain and withoutsuffering; however, this is subjective and priorities are based on personal,cultural, social and religious perspectives. Currently, there is insufficientdata to generalise rural residents’ perspectives and what it means forthem to die well. Given the extreme importance of a ‘good death’, thereis a need for further studies to elicit rural patient and family caregiverperspectives.
Keywords: dying well, end-of-life, good death, palliative care, rural, scopingreview
Introduction
Death is an individualised experience with cultural,religious and political values and beliefs influencingthe quality of dying. (Cottrell & Duggleby 2016). The‘good death’ is one of the main objectives, if not the
sole aim, of palliative care. It is a dynamic conceptwhose meaning has changed over time in line withurbanisation and advances in medical technologies(Aries 1974, Kellehear 2008). Some argue the contem-porary notion of the ‘good death’, fostered by thehospice/palliative movement, is idealised and ‘limits
© 2016 John Wiley & Sons Ltd 1
Health and Social Care in the Community (2016) doi: 10.1111/hsc.12385
spontaneity’ (Cottrell & Duggleby 2016). Historically,death and dying have moved from an open publicactivity at home to a more private and institution-alised process, hidden away in hospitals or otherinstitutions such as nursing homes (Gomes & Higgin-son 2008), despite studies increasingly indicating thehome as the preferred place (Gomes et al. 2013, Rains-ford et al. 2016).
The common essential features of the modern‘good death’ are thought to include being pain free,maintaining dignity, support of family, autonomy indecision-making for the dying person and opportu-nity to ‘sort out personal affairs’ (Raisio et al. 2015,Davies et al. 2016, Meier et al. 2016). However, differ-ent stakeholders have different priorities and expecta-tions of end-of-life care (EoLC) (Holdsworth 2015,Davies et al. 2016, Meier et al. 2016). Ideally, it mustreflect the needs of the community and in particularthose of the dying person and their carers (Cottrell &Duggleby 2016). One voice that remains largelyunheard is that of rural palliative care patients andtheir families (Bakitas et al. 2015).
One challenge in finding the rural voice is thedifficulty in establishing consensus on a ‘rural’ defi-nition. While efforts are made to define ‘rural’ geo-graphically (based on population density ordistance from services), Wilson et al. (2009a,b) sug-gest rural people define themselves as ‘rural’ andperceive themselves as different from urbanites.Studying rural views of the ‘good death’ is impor-tant because these may differ from urban views(Spice et al. 2012). In addition, ageing is more pro-nounced in rural areas, and rural areas typicallyhave fewer health and social services as comparedto urban areas (Downing & Jack 2012). For exam-ple, the growth in Australia’s population aged65 years or older is expected to rise by 139%between 2000 and 2030, with a 180% increase antic-ipated in rural areas (Australian Report 2008).
The purpose of this scoping review was to iden-tify and chart current research knowledge on the‘good’ rural death through the perspectives of ruralresidents, including rural patients with a life-limit-ing illness, rural family members/informal care-givers (FCG), rural healthcare providers (HCP) andthe wider rural community; and to identify gaps inthe literature for areas for future studies to betterunderstand the concept of ‘good death’ for ruralpeople. This review will add to an understandingof the rural perspective on the ‘good death’ toguide researchers, healthcare professionals and pol-icy makers in future planning and development ofrural EoLC.
Methods
Protocol
Currently, there is no standardised definition ormethodology for scoping reviews (Peters et al. 2015);however, the definition commonly applied is thatused by Arksey and O’Malley (2005). The protocolused in this review is based on the methodologicalframework first described by Arksey and O’Malley(2005), enhanced by Levac et al. (2010), Colquhounet al. (2014) and Daudt et al. (2013) and later refinedby the Joanna Briggs Institute (JBI 2015). The original(Arksey & O’Malley 2005) and modified (Levac et al.2010) frameworks consist of six stages (JBI, 2015).
Despite scoping reviews becoming increasinglypopular (Tricco et al. 2016), there are no standardisedreporting guidelines. Tricco et al. (2016) identified theneed for guidelines to ensure scoping reviews are vali-date and reliable. The reporting of this review is basedon the 2016 scoping review published by Tricco et al.(2016), the team that is developing the standardisedreporting guidelines, as well as the PRISMA-ScopingReview (PRISMA-ScR) guidelines (Equator Network,2016). While quality assessment is not a requirementof the JBI guidelines, it will be included in this review(Levac et al. 2010, Daudt et al. 2013).
Eligibility criteria
Scoping reviews have a broad approach and includeany existing literature regardless of study design, dis-cipline or quality. In order to answer the researchquestion ‘How do rural residents describe the “gooddeath” concept within a rural setting?’ eligibility crite-ria were developed to clearly identify the purpose ofthe review and guide the reviewers in deciding whicharticles to include.
The eligibility criteria were developed using theJBI (2015) guidelines using the Participants, Conceptand Context acronym.• Participants: Rural residents including rural
patients with a life-limiting illness, rural FCGs,rural HCPs and the wider rural community asthese are the most appropriate to provide the ruralperspective. No age filter was applied.
• Concept: The principal concept under review wasthe ‘good death’ in a rural setting as describedthrough the personal experiences or perspectivesof rural residents; collected by interviews, surveysor extensive field work observations. The term‘good death’ was either used explicitly or implied.
• Context: Rural or remote; all countries and territo-ries were considered; no standardised definition of
© 2016 John Wiley & Sons Ltd2
S. Rainsford et al.
rurality was used. Articles that included urbanand rural data were considered providing therural data were clearly identifiable.
Information sources and search strategy
The principal researcher conducted the comprehensiveliterature search. Five electronic databases (PubMed,CINAHL, Scopus, PsycINFO and Web of Science) weresearched from 2 January to 14 February 2016. The fol-lowing keywords and Medical Subject Heading(MeSH) terms were used in the final search: (“gooddeath” OR “managed death” OR “good enough death”OR “tamed death” OR “dying well” OR “peacefuldeath”) AND (Rural OR Remote). This iterative processhad ‘peaceful death’ omitted from the original search.All study designs were included; no date filter wasapplied; only English language papers were included.
The initial searches identified 377 articles. Thesewere downloaded to ENDNOTE X7, merged andduplicates were deleted (338 articles). The referencelists of all retained articles were scanned for additionalstudies. The principal author also hand-searched recentissues (July 2014–Jan 2016) of eight relevant journals(Palliative Medicine, Journal of Palliative Medicine,Palliative and Supportive Care, Australian Journal ofRural Health, Journal of Rural Health, Social Scienceand Medicine, Health and Place, Death and Dying) foradditional articles. The Cochrane Library, CareSearchdatabase and OpenGrey repository were searched forgrey research literature. The authors of three studiesreporting mixed rural/urban geographical data werecontacted; however, rural data were confirmed as notspecified, and so these three articles were excludedfrom this review.
Study selection process
Phase 1 screening (review of titles) was carried outindependently by two authors. At least two authorsindependently screened the abstracts and/or full textof potential articles (see Figure 1 for flow chart).Rejected articles were either clearly irrelevant or thosethat addressed the topic in general but failed in oneor more of the inclusion criteria. Due to geographicaldistances, differences were discussed via email andresolved by consensus.
Data extraction and charting
Data from each accepted article were extracted by theprincipal author and entered into a table according topredefined headings (JBI, 2015). Three co-authors
independently read the retained articles and addedtheir comments to the table indicating agreement/dis-agreement/additions/deletions. Differences were dis-cussed via email. The final version was emailed to allco-authors for verification.
Assessment of quality
The current scoping review methodological guideli-nes do not require a formal quality assessment of eli-gible articles to ensure valuable insights reported inlower quality studies are not excluded. However, aquality assessment was conducted in this review toassist in validating the quality of the literature. Alleligible articles are retained in this scoping reviewregardless of their quality.
The quality of each study was assessed indepen-dently by four co-authors. Some high-quality studiesreceived a lower score as the assessment was basedon aspects relevant to the rural ‘good death’ and notof the study per se. All studies were rated to be oflow, medium or high quality based on a simple scor-ing system described by Gomes et al. (2013) andmodified to account for the rural ‘good death’ focus.Two additional items were included: (1) clarity ofrural definition and (2) validity of informant(prospective = 2, retrospective = 1, well commu-nity = 0).
Synthesis
The synthesis included both a quantitative analy-sis of the actual scoping review and a qualitativeanalysis of the content of the included articles.Both analyses were conducted by the principalauthor and verified by one of three co-authors.The quantitative analysis is charted and describedin frequencies. A thematic analysis of the qualita-tive content was conducted by downloading theeligible articles into NVivo-10, coding for majorthemes and reported narratively. Due to hetero-geneity within a small number of studies, a meta-analysis and analysis according to informantswere not possible. Some informant groups hadonly one study identified.
Results
Literature search
A total of 377 potential articles were identified fromthe electronic searches. After merging the searchesand removing duplicates, 338 manuscripts were
© 2016 John Wiley & Sons Ltd 3
Rural residents’ perspectives on the rural ‘good death’
identified and 13 remained after two screeningphases (see Figure 1). Hand-searches of journals andreference lists identified an additional seven articles.Consequently, 20 articles reporting on 17 studies andone systematic review were included in this scopingreview (see Table 1).
Study characteristics
The included articles were published 2000 through2015 (see Table 2), with data collected between 1981and 2013. Rural perspectives were reported in articlesfrom both developed and developing countries –
Canada (5), USA (4), Kenya (2), United Kingdom (2)and one each from Norway, Vietnam, Bangladesh,Papua New Guinea, Uganda, Malawi, Ghana andTanzania. One study (Grant et al. 2011) and the sys-tematic review (Gysels et al. 2011) collected data frommore than one country in sub-Saharan Africa. OneCanadian manuscript (Wilson et al. 2009b) compareddata collected from two studies (Wilson et al. 2009a,Viellette et al. 2010) conducted in different Canadianprovinces using the same study protocol. Withthe exception of two articles (Gysels et al. 2011,Knight 2014), all studies were described as rural; noremote communities were specified. No standardised
Articles after searches merged, duplicates removed (n = 338)
Phase 1 screening: Titles reviewed by SR and RM
Irrelevant (n = 290)
Phase 2 screening: Abstract and/or full-text articles assessed for
eligibility by SR and RM or NG orDW or CP or RW (n = 48)
Articles excluded (n = 35)• Not good death (16)• Not rural (2)• Rural data not identified (4)• Not relevant (4)• Duplicate [poster] (2)• Workshop abstract (1)• No personal perspectives (6)
Eligible studies (n = 13)
Studies added through hand search of references and journals (n = 7); Grey literature search (n = 0)
Articles included in scoping review (n = 19) reporting 17 separate
studies Systematic
review (n = 1)
Articles identified through database searching (n = 377)
Figure 1 Flow diagram of scoping review selection process including reasons for exclusion.
Source: Modified flow chart as described by Moher et al. (2009).
© 2016 John Wiley & Sons Ltd4
S. Rainsford et al.
Table
1Summary
ofincludedstudies
Purpose
Inform
ants
Ruraldefin
ition
Methodology
Gooddeath
description
Findingsin
relationto
gooddeath
Quality/
Lim
itations/
Bias
Arnaertetal.
(2009)
Western
Quebec
(Canada)
RuralN
=5
Toexplore
home-care
nurses’attitudes
towardspalliativecare
(PC)
inaruralcommunity.
Palliativecare
journal
5home-care
nurses;all
residents
of
therural
community.
Vast,sparsely
populatedarea
geographically
isolatedwith
few
healthor
socialservices.
Ruralstudy
Qualitativeface-to-
facesemi-
structuredITVs
andFGD;
thematicanalysis.
‘Ifthey[thepatie
nts]die
withdignityandthey
are
consciousrightto
theend,andthey
haven’tsuffered,
then...theyhavehad
agooddeath’[nurse].
Six
themesidentifie
d:
1Community
embeddedness;
2Maintainingpatients’
QoL
3Multidisciplinary
collaboration
4Emotio
naland
physicalisolation
5Needforprofes
sionaldevelopment
6Organisational
issues.
Medium
Smallsample
size;
employedbyPC
service.
Beckstrand
etal.(2012,
2015)
Wyoming,
Utah,
Nevada,
Alaska
(USA)
RuralN
=236
Todiscoversize,
frequency,
and
magnitudeof
obstaclesin
providing
end-of-lifecare
(EoLC)
inruralemergency
departments
(ED).
Emergencynursing
journal
52/73rural
hospitals
agreedto
participate;
236/508
surveys
returnedby
registered
nurses.
Smallrural
communities
withcritical
access
hospitals
(CAH).
Ruralstudy.
Cross-sectional
surveydesign;57-
item
survey
included15
demographic,3
openendedand
39Likert
questions.
Tobecomfortable
and
freeofpain,thefamily
isable
toperform
EoL
rituals,others
are
respectfulofthe
patient’s
dignity,and
familieshave
adequate
timeto
say
goodbye.
GD
=idealdeath
Obstaclesto
ideal
death
inruralED
includespersonally
knowingthepatientor
family;lackofpriva
cy
forthedyingpatient
dueto
poordesign;
issueswithfamily
members;unknown
patientwishes;nurse/
physicianpower
struggle;Phonecalls
from
family,etc.
detractfrom
timewith
patient.
Medium
Conveniencesample;
ED
nurses;focus
mainly
suddenor
accidentaldeaths
within
theED.
Biggs(2014)
Canada
RuralN
=19
Toexplore
how
dying
personsandtheir
caregivers
experience
spiritualcare
intheir
homesasanaid
to
dyingwellandhaving
aGD.Toexamine
effectofrural
geographyonthe
experience.
Doctoralthesis
10dying
personsand9
caregivers.
GreyandBruce
Counties
describedas
rural
comprising
townsof
varioussizes
andfarm
s(no
specific
details).
Ruralstudy.
Phenomenological
studyand
interpretive
analysis;semi-
structuredITVs.
Withspiritualcare
‘people’s
needsare
fulfilled,theirsuffering
islessenedandthere
isanincreased
possibility
ofthem
experiencingagood
death’.
EoLspiritualcare
facilitates‘GD’by
helpingdyingpersons
andtheirFCG
to
experienceconnection
andsupportto
people
andtraditionsthey
considernecessary
to
die
well;
provides
personalinner
resourcesincluding
hope,comfort,self-
worth,strength
to
cope,andpeaceof
mind.
Ruralresidency
contributedpositively
totheirexperienceofa
GD.
High
Smallsample
sizelim
itationis
counterbalance
dbydepth
of
interviews.
© 2016 John Wiley & Sons Ltd 5
Rural residents’ perspectives on the rural ‘good death’
Table
1(continued)
Purpose
Inform
ants
Ruraldefinition
Methodology
Gooddeath
description
Findingsin
relationto
gooddeath
Quality/
Lim
itations/
Bias
Cruickshank
etal.(2010)
Scotland
RuralN
=24
Tounderstandthe
impact
onpatientcare
byuseofsyringe
drivers
(SD)in
rural
communities
Palliative
care
journal
4term
inal
patients,8
inform
al
carers,12
community
nurses.
‘TypicalScottish
rural
community’.
Town15,000
and
surrounding
ruralarea.
Ruralstudy.
Phenomenological
approach;Patient
ITVsathome;
carer
bereavement
ITVs;2community
nurseFGDs.
Goodpain
reliefmakes
aGD;however,many
otherissuesto
consider‘it
[syringe
drive
r]helpsthepain
butthat’sallithelps’
[patient].
Factors
influencing
gooddeath:
1Goodanticipatory
care
2Sym
ptom
management
(physical,
psychological)
3Open/honest
communication
–family
and
patie
nt
EoLC
iscomplex;
Communitynurseis
centralto
EoLC;
Ruralchallenges
includedistances
andlim
itedstaff.
Medium
Patients
enrolledin
palliativecare
service;
hadSD
atrecruitment;
biasedtowards
symptom
control.
NotfocusedonGD;
focusedon
determ
iningEoLC
needsin
arural
community
Counts
andCounts
(2004)
WestNew
Britain
(PapuaNew
Guinea)
RuralN
=whole
community
‘Thegood,thebad,and
theunresolveddeath
inKaliai’
Socialscienceand
medicinejournal
Authors’
observatio
ns
andinterviews
withresidents.
IsolatedNW
island;1000
Melanesian
Lusi-Kaliai
residents;
accessible
only
bysea.
Government
clinic
operated
atRoman
Catholic
Mission.
Ruralstudy.
Qualitative
Long-term
Anthropological
field
work
(1966–
1967,1975–1
976,
1981,1985).
Casestudies
‘Agooddeath
isunder
thecontrolofthedying
personandis
the
resu
ltofthenatural
proce
ssofageing’.
[author];‘death
is
eithergoodorbad
dependingonwhether
itis
theconsequence
ofbadsocial
relatio
nshipsand
causessocial
upheaval’.
GD
=‘peacefuldeath’;
5elements:social(at
peacewithothers,
mutually);
psychologicalor
spiritual(atpeacewith
one’s
ownlifeand
death
and‘soul’);
timing(naturaldeath/
old
age);spatial(at
home,surroundedby
relativesA
GD
rarely
happens;itis
acontrolled,
quietdeath
ofelderlyas
resultofthenatural
processofageing.
Baddeath
impliesa
rupture
ofsocial
relationsandresultsin
thedestructionof
peaceandsocial
order.
Combinatio
nof
traditionaland
Christianbeliefs.
Socialdeath
versus
physicaldeath
Low
Observation–
researcherbias
howeverdata
collectedovermany
years.
Understandingofgood
andbaddeath
isvery
muchbasedon
localisedcultural
constructionsandmay
notnecessarily
reflect
generalbeliefs
and
customs.
Old
study–findings
maynolongerbe
relevant.
© 2016 John Wiley & Sons Ltd6
S. Rainsford et al.
Table
1(continued)
Purpose
Inform
ants
Ruraldefinition
Methodology
Gooddeath
descriptio
n
Findingsin
relationto
gooddeath
Quality/
Lim
itations/
Bias
Devik
etal.
(2013)
RuralN
=5
Norw
ay
Toexplore
and
understandlived
experienceofolder
cancerpatients
living
alone.
Oncologynursing
journal
5olderpatients
withincurable
cancer;
>70years;
livingontheir
own.
Communities
with
low-population
density;
population920
to7775.
Ruralstudy.
Qualitative;
phenomenological
herm
eneutical
approach;
narrative
interviews.
‘Gooddeath’is
implied;
could
beinferredto
be
anoble
death;
acceptancecomes
throughmaintaininga
willto
live;thehopeto
stayaliveis
notdeath-
renouncing,butrather
life-affirm
ing.
Importanceof
maintainingdignity,
identity
andvalue;
enduringbykeeping
hopealiveby
continuing
chemotherapyasa
meansofdelaying
death
andretaininga
positiveoutlook;
maintainingautonomy
bymobilisingpersonal
resources;notto
becomeaburdenon
family,friendsand
services.
Challenges:coping
withconflictingfeelings
ofhopeanddespair;
navigatingaloneas
patientandself-care
provider;livingupto
expectationsofbeing
agoodpatient;lim
ited
control;riskoflosing
one’s
identityand
value;lossofform
er
life;servicesnot
alwaysavailable
in
ruralareastherefore
needto
beself-reliant.
Medium
Smallsample;
recruitmentbias
(polyclinic);patients
undergoing
chemotherapy;
patie
nts
livingalone.
© 2016 John Wiley & Sons Ltd 7
Rural residents’ perspectives on the rural ‘good death’
Table
1(continued)
Purpose
Inform
ants
Ruraldefinition
Methodology
Gooddeath
descriptio
n
Findingsin
relationto
gooddeath
Quality/
Lim
itations/
Bias
Dilger(2008)
Tanzania
Toexplore
how
moral
perceptionsofHIV/
AIDS-relatedillness
anddeath
inrural
Tanzania
are
related
tosocialandcultural
practicesofdisease
interpretation,patie
nt
caringandburial.
AnthropologyJournal
Youngmen
andwomen
withHIV/AIDS
(minorethnic
group.)
Mara
Regionin
Tanzania;
rural;borders
onLake
Victoriain
the
westand
Kenyain
the
north;farm
ing
andfishing.
Ruralstudy.
Anthropological
field
work;
Research1999–
2000.
Acontrolled,quiet
death
ofelderlyas
resultofthenatural
processofageing.
‘Baddeath’–physical
andspiritualwith
social,culturaland
moralcomponents.
Suicideis
anexample
ofabaddeath
?negationandrejection
ofsocialorder;HIV/
AIDS
isa‘bad
disease’inferringa
‘baddeath’-painfulfor
individuals
andtheir
families;premature
death
breakscontinuity
ofwhole
familiesand
lineagesasyoung
people
die
without
leavingchildrenfrom
‘legitim
ate
relationships’;disturbs
socialequilibrium
and
kinship
networks;
considereda‘final
death’asprevents
fulfillingofritual
requirements.
Low
Field
work
andcase
studieswithno
reportingof
methodology;no
refle
xivity;Resultspossibly
notgeneralisable
dueto
small
ethnic
groupwithspecific
culturalpracticesandyoung
ageofinform
ants..
© 2016 John Wiley & Sons Ltd8
S. Rainsford et al.
Table
1(continued)
Purpose
Inform
ants
Ruraldefinition
Methodology
Gooddeath
descriptio
n
Findingsin
relationto
gooddeath
Quality/
Lim
itations/
Bias
Easom
etal.
(2006)
USA
RuralN
=9
Pilotstudyto
evaluate
effects
ofan
education
interventiononEoLC
knowledgeand
perceptionsin
GD
definition.
Palliativecare
journal
Convenience
sample;9rural
nursesworkingin
assistedlivingand
nursinghome
environments
Rural=access;
distance
>75milesfrom
large
metropolitan
area.
Ruralstudy.
Mixedstudy;written
questionnaire.
Inregard
GD
attitudes:Qualitative
:
2open-ended
questions.
Q1:‘W
hat
isyourbelief
regardingwhat
constitutesa“good
death”?’
Q2:‘W
hatim
pactdo
culturalvalueshave
ondeath
anddying?’
Todie
painlessly,
comfortably
and
hopefully
inone’s
sleep.
Attitudesand
perceptionsof
participants
changed
indefiningwhat
constitutesa‘good
death’.
Q1:Pre-education:
todie
painlessly,
comfortably,and,
hopefully
insleep.
Post-education:
addingim
portanceof
family
involvement;
Respectand
compliancewith
patientandfamily
wishes.
Q2:Nochangein
knowledgeand
attitudesrelatingto
significanceofcultural
valuesonthe
proce
ssofdeath
anddying.
•Communication.
•Cultu
ral
competency.
Low
Smallconvenience
sample;Recruitment
bias.
Pilotstudy.
NotfocusedonGD-
evaluatedstaffEoLC
educatio
n
programme.
Feltetal.
(2000)
Kansas,
USA
RuralN
=approx.50
Toaddressunder-
utilisationofrural
hospiceservices;To
determ
ineattitudes,
values,beliefs
and
practices
surroundingEoLC.
Palliativecare
journal
Targeted
communities;
purposiveselection
ofactive
community
leaders.
Six
rural
communities
inKansas.
Ruralstudy.
Qualitative;6FGDs
of7to
10people;
discussions
recorded,
transcribed,reviewed
individually;
comparisonsmade
betweengroups?
themesidentifie
d.
Bestscenariofor
one’s
owndeath
‘in
mysleep’or‘hopeit
happensquick’;at
homeor‘someplace
comfortable,quiet,
andpeaceful,with
family’andloved
onesnearby;to
‘makepeace’,to
‘forgiveandforget’,
tobehonestabout
whatwasreally
happening;
perm
issionto
die
from
lovedones
very
importantto
a
peacefuldeath.
GD
=‘bestscenario
forowndeath’
Importantthemes:
1Freedom
and
independence
2Honesty
andtrust
3Im
portanceof
family
4Therightto
inform
atio
n
5Participants
expressedmore
fearoverthe
mannerofdeath
thandeath
itself;
fearatechnological
death
asopposed
toagooddeath.
Somewantedto
die
aloneafter
closure
with
lovedones.
Medium
Selectio
nbias;
minority
groupsnot
represented;script-
guideddiscussion.
Focusedon
determ
iningEoLC
needsin
arural
communityrather
thanGD.
© 2016 John Wiley & Sons Ltd 9
Rural residents’ perspectives on the rural ‘good death’
Table
1(continued)
Purpose
Inform
ants
Ruraldefinition
Methodology
Gooddeath
description
Findingsin
relationto
gooddeath
Quality/
Lim
itations/
Bias
Grantetal.
(2003)
RuralKenya,
sub-SaharanAfrica
RuralN
=56
Todescribethe
experiencesand
needsofdying
hosp
italpatients
and
theirFCG
to
determ
inewhat
constitutesaGD
in
sub-SaharanAfrica.
Palliative
care
journal
32adultpatients
(cance
rorAIDS);
24carers;mostly
Christianhowever
traditionalbeliefs
stillstrong;
identifie
dby
hospita
ldoctor;
stratifie
dbyhome
andhospitalcare.
Eastern
slopes
ofMtKenya
(Meru
District).
Population
density363/
km
2to
65/km
2-
poverty,
drought,poor
infrastructure.
Population
mostly
Christian.
Ruralstudy.
Qualitative
;Semi-
structuredITVs;not
recorded,hand
writtennotes+
impressionsand
observations;
transcultural
researcher;
multidisciplinary
advisory
group;
multiple
data
sources;ITVs
completiononce
saturationachieved;
constantcomparison;
thematicanalysis.
Dyingwithdignityor
dyingwellis
outside
thegraspofmost
dueto
overw
helm
ing
pain,poverty,sense
ofburden,guilt,and
needforbasicsof
life.
Positive:supportof
closefamily
relationships,care
shownbycommunity
andreligious
fellowshipshelped
meetmanyoftheir
emotio
nal,social,
andspiritualneeds;
acceptedby
relative
s.
Negative:physical
needswere
often
unmet.Unresolved
suffering-pain,
poverty,guilt
ofusing
allavailable
family
resourcesto
payfor
treatm
entandcare.
Needs:
accessible
pain
relief,affordable
clinic
orinpatient
care,help
tocope
withtheburdenof
care.
Holisticapproachto
theGD.
High
Lim
ited
generalisability
to
sub-SaharanAfrica.
Selectionbias-
recruitedfrom
hospital.
Smallsample,single
ethnic
group;
Interviewsnot
recorded,however,
rigourin
recall;
cancer/AIDS
focus.
Grantetal.
(2011)
Uganda,
Kenyaand
Malawi
RuralN
=162
Todescribepatie
nt,
family
andlocal
community
perspectivesonthe
impactofthree
communitybased
PC
interventionsin
sub-SaharanAfrica.
Palliative
care
journal
33patients
with
advancedillness,
27family
carers,
36staff,25
volunteers,29
communityleaders
andobserved
clinicalcare
of12
patients.
Uganda:mobile
PC
service
120km
from
Kampala;
includesfishing
communities
onLakeVictoria.
Kenya:rural,
240km
from
Nairobi.
Malawi:peri-
urbandistricts
ofBlantyre.
Influenceof
Christian
missionaries.
Ruralstudy.
Rapid
assessment
evaluatio
n;
Qualitative
;
photographic
ethnography;rapid
evaluatio
nfield
studies;ITVswith
keyinform
ants;
directobservationsof
clinicalencounters;
review
oflocalPC
inform
ation;data
triangulation;
thematicanalysis.
‘Toletpeople
die
in
peace’;‘now
patients
die
withdignity’with
a‘senseofhope’.
Gooddeath
implied
Gooddeath
facilitated
by:Patients
and
family
carers:Pain
control;beingtreated
withdignityand
respect;being
supportedathome
reducesphysical,
emotio
naland
financialburdenof
travel;practical
support;rapid
access
toclinicalandsocial
supportnetworks.
GD
requiresa
holisticapproach
especially
goodpain
management.
Medium
Serviceevaluation;
selectionbias;
small
samples.
Palliative
care
present?
may
inhibithonest
responses.
S. Rainsford et al.
© 2016 John Wiley & Sons Ltd10
Table
1(continued)
Purpose
Inform
ants
Ruraldefinition
Methodology
Gooddeath
descriptio
n
Findingsin
relationto
gooddeath
Quality/
Lim
itations/
Bias
Gysels
etal.
(2011)
Sub-SaharanAfrica
Tosynthesise
qualitative
research
onEoLto
inform
policy,practiceand
furtherresearch.
Palliativecare
journal
–Studiesrelating
totheGD
were
conductedin
ruralGhana,
Tanzania,and
Kenya.
Systematicreview
Ghana:Baddeath
is
adeath
thatcome
tooearly,orbelieved
tobepunishmentfor
sins.
Tanzania,Kenya:
baddeath
relatedto
‘baddisease’(H
IV/
AIDs).
GD
versusbaddeath
andstigma.Cultural
beliefs
impactEoLC
forpatient,family
andcommunity.
Ghana:elderly,
death
welcomed,
bringspeaceand
rest.Baddeaths–
early,punishmentfor
sin.Agoodfuneral
waspartofagood
death.
Tanzania:studies
relatedto
stigma
attachedto
HIV/
AIDs,associated
withim
moral
behaviour=bad
death.
KenyaandTanzania:
HIV/AIDsdeath
deniedstatusof
ancestors
?threatenedcontinuity
ofclanand
community;buried
outsidethe
homeland?
lossof
dignityand
complicated
bereavement.
Systematic
review;
methodologically
soundnotGD
focused.
Systematicreview
base
donEoLC
in
sub-SaharanAfrica;
predominantlyHIV/
AIDSamongyoung
people;Findingsnot
generalisable
outside
this
cohort.
© 2016 John Wiley & Sons Ltd 11
Rural residents’ perspectives on the rural ‘good death’
Table
1(continued)
Purpose
Inform
ants
Ruraldefinition
Methodology
Gooddeath
description
Findingsin
relationto
gooddeath
Quality/
Lim
itations/
Bias
Huyetal.
(2007)
Vietnam
Rural
N=approx.
80.
Toexplore
sociocultural+
healthsystem
factors
impacting
registeringofdeaths
bylaypeople.
Worldhealthand
populationjournal
Purposiveselection;
Farm
ers,livedin
hamlet>5years.
Commune6000
–10,000;each
dividedinto
a
numberof
smallhamlets.
Vietnamese
folk
religion
and
Christianity.
Ruralstudy
Qualitative
9FGDs
(6–1
1participants);
audio
recordedand
transcribed;
translatedin
to
English;thematic
analysis
GD
isan‘honourable
death’
Typesofdeath:
•Elderly(natural
processand
deserved)
•Young(w
orking
adults):
GD
=‘honourable
death’;‘Bad
death’=death
of
dishonourable
personwhohad
livedabadlifeor
stigmatise
ddisease
(HIV
orsuicide)
•Youngdeath
(schoolaged)
•Infantdeath
(either
1or12years)
Badandinfant
deathsmore
likely
tobeunreported
dueto
family
stigma
andbabiesnot
recognisedasa
‘person’
Low
FGDspossibly
not
appropriate
inthis
culture
andmay
inhibitdiscussion;not
generalisable;no
reportingof
researcherbiasor
reflexivity;no
indicatio
nof
participants’personal
experiencewith
dying(implied).
Joarderetal.
(2014)
Bangladesh
RuralN
=8
Toexplore
perceptionsofthe
meaningofdeath
amongelderly
people
ina
Bangladeshi
community,andto
understandhow
the
meaningofdeath
affects
one’s
overall
well-being.
Cross-cultural
gerontologyjournal
5male
+3female
elderlyvillagers.
Data
collected
2008.
Ruralvillage
(Kakabo),
Savarsub-
district25km
from
Dhaka.
MostlyMuslim
;
few
Hinduand
Christian.
Ruralstudy.
Qualitative
;purposive
sampling;in-depth
ITVs;daily
routines;
inform
aldiscussions;
Participatory
Rapid
appraisaltoolbox.
Codingand
categorisingpatterns.
Peacefully,without
suffering,
surroundedbyfamily
Physicalversus
spiritualdeath;
dualityofbodyand
soul.
‘Gooddeath’–
peacefully,without
suffering,surrounded
byfamily.This
mannerofdeath
was
givenbyGod;the
resu
ltofgood
actionsin
life.
‘Baddeath’=caused
byhanging,
poisoning,homicide,
andaccidents;
earnedthroughbad
deed;bad‘karm
a’.
Medium
Shorttimeinterval
(3days);age
difference
between
interviewers
and
participants
(respondents
considered
interviewers
too
na€ıveto
discuss
seriousissues);
beliefs
base
don
localcultural
traditionswith
religiousinfluences
limiting
generalisability.
S. Rainsford et al.
© 2016 John Wiley & Sons Ltd12
Table
1(continued)
Purpose
Inform
ants
Ruraldefinition
Methodology
Gooddeath
description
Findingsin
relationto
gooddeath
Quality/
Lim
itations/
Bias
Knight(2014)
England
RuralN
=4
Toexplore
theviews
ofclergyaboutthe
constituentelements
ofa‘gooddeath’.
4Churchof
Englandclergy
workingin
ruralor
townparishes.
(10clergy
including3rural,1
town,2suburban;
4notspecified)
Dioceseof
Worcester.
Urban/Rural
study;rural
data
identifie
d.
Qualitative;random
sample;structured
ITVsusingflash
cards;thematic
analysis
andscoring
ofcards.
Differentmeaningsof
‘GoodDeath’
dependingon
perspectives,
cultures;individual
concept.
Spiritualcare
slightly
more
significantthan
emotionalcare
and
more
importantthan
physicalcare.Social
domain
lowest
priority.
Spiritualcare:‘tobe
atpeacewithGod’;
‘tosaygoodbyeto
importantpeople
in
mylife’
Emotionalcare:
‘dignity;senseof
humour’
Physicalcare:pain
free;‘Tohavehuman
touch’.
Socialcare:to
have
myfamily
withme’
•Keepingsenseof
humour+family
involvementhighly
valued;Tosay
goodbyemore
highly
ratedthan
dyingathome.
•Peacefulnessfor
HCP
=absence
of
distressing
symptoms;for
theclergy=
Maintainingsense
ofhumour,senseof
dignity;
beingat
peacewithGod.
Medium
Smallsample
size;
clergybias;single
rese
archer.
© 2016 John Wiley & Sons Ltd 13
Rural residents’ perspectives on the rural ‘good death’
Table
1(continued)
Purpose
Inform
ants
Ruraldefinition
Methodology
Gooddeath
description
Findingsin
relationto
gooddeath
Quality/
Lim
itations/
Bias
vander
Geest
(2004)
Ghana
RuralN
=35
Anessayconsidering
‘gooddeath’and
‘baddeath’in
rural
Ghana.
Socialscienceand
medicinejournal.
Residents,mostly
farm
ers,ofKwahu-
Tafo,southern
Ghana.
Aruraltown;
population
6000;farm
ers
andtraders.
Christian,
mostlyRoman
Catholic.
Ruralstudy.
Anthropological
fieldwork
carriedout
interm
ittentlyfrom
1994to
2004.
Interviews,
participant
obse
rvations,school
questio
nnaires,
FGD.
Agoodandpeaceful
death
comes
‘naturally’aftera
longandwellspent
life.
‘Gooddeath’=a
peacefuldeath:
havingfinishedall
business,made
peacewithothers;
beingatpeacewith
owndeath;notby
violence,accident,
fearsomedisease,
withoutmuchpain;at
home,surroundedby
childrenand
grandchildren;a
death
acceptedby
therelative
s.
5aspects:social,
psychological,
spiritual,timing,
spatial.
However,‘good
death’is
ambiguous
asqualityis
opento
interpretatio
nbyall
stakeholders.
Medium
Notastudywith
methodsbut
extensivefield
work,
opento
researcher
bias.Noreflexivity
provided;not
generalisable;ethics
notdescribed.
S. Rainsford et al.
© 2016 John Wiley & Sons Ltd14
Table
1(continued)
Purpose
Inform
ants
Ruraldefinition
Methodology
Gooddeath
descriptio
n
Findingsin
relationto
gooddeath
Quality/
Lim
itations/
Bias
Vielletteetal.
(2010)
Quebec,
Canada
RuralN
=36
(conce
rning
70cases)
Toobtain
perspectiveson
whatconstitutesa
gooddeath
from
personslivingin
Frenchspeaking
ruralQuebec.
Palliativecare
journal
ITV:25long-term
ruralcommunity
members
with
personalor
professional
experienceof
death
anddying.
FGD:11
healthcare
providers,
community
servicesandEoL
policymakers/
administrators.
RegionA:
considerable
distancefrom
Quebeccity.
RegionB:
closeto
QuebecCity.
Ruralstudy.
Qualitative
ethnography;
validationofthemes
previously
identified
(Wilson);25ITVs;2
FGD(5
+6);audio
taped;transcribed;
coding;themes;
iterative;saturation.
‘Enjoyingagood
qualityoflifewhile
dying(andtheir
family)wasessential
toagooddeath’.
GD:fourdim
ensions–
physical,spiritual,
socialandemotional
–psychological.
Essentialelements:
1Controlover
physicalpain
2Notbeinga
burdenonfamily
3Ability
tomake
choicesand
controloverlife
4Beingin
one’s
environment,
surroundedby
family
+friends
5Beingspiritually
accompanied
6Good
communication
7Goodrelationsh
ips
withcare
and
serviceproviders.
RuralSpecific:to
die
inone’s
own
community,
relianceon
neighboursupport.
Somewould
prefer
tohavelesscare
andfewerservices
thanberemoved
from
community
(family
and
friends).
Gapsin
available
servicesare
limitingfactors
toaGD.
High
Studyin
2small
areasofalarge
province
;
retrospectiveviews-
someofmanyyears;
noFirstNations;few
males;recruitment
bias:
self-selection.
© 2016 John Wiley & Sons Ltd 15
Rural residents’ perspectives on the rural ‘good death’
Table
1(continued)
Purpose
Inform
ants
Ruraldefin
ition
Methodology
Gooddeath
description
Findingsin
relationto
gooddeath
Quality/
Lim
itations/
Bias
Wilsonetal.
(2009a,b)
Alberta,
Canada
RuralN
=34
Togain
aconceptual
understandingofthe
‘gooddeath’in
Englishspeaking
ruralAlberta.
Palliativecare
journal
ITVs:13community
members
with
personalor
professional
experienceof
death
anddying;
self-selection.
FGD:21
healthcare
providers
and
policymakers.
RegionA:
Metro-
adjacent;
RegionB:
Non-m
etro-
adjacent.
Ruralstudy.
Qualitative
ethnography;
Documentreview;
siteobse
rvations;
interviews(13);
themesvalidatedby
2focusgroups
(9+12);audio
tapes
transcribed;coding;
themes;iterative;
saturation.
Thegoodruraldeath
is
contextualised.Dying
outsideone’s
rural
communityis
considered
a’baddeath
Fourthemes:
1‘Ruralpeople
are
adistinctcultural
group’withunique
perspectiveson
EoLC
andGD;
ruralpersonsneed
tobein
theirhome
communitywhen
dying.
2Deepconcern
aboutcommunity
anditsmembers;
desireto
provide
highly
individu
alisedcare
aim
ed
atmaintainingthe
dignityand
addressingthe
self-directedcare
preferencesofthe
dyingperson.
3Loosenetwork
of
carers
(paid
and
volunteer)
4Uniquechallenges
indevelopingand
maintainingEoL
services
Ruralpeople
have
lessaccessto
specialistpalliative
care.
High
Lim
itationsnotnoted
howeverresearcher
reflexivityandbiasis
acknowledged;
selectionbias;small
sample.
Findingsreportedare
inrelationto
gooddeath
anddonotnecessarily
reflectfullfindingsofthestudy.
ABBREVIATIO
NS:PC
palliativecare;ITVinterviews;FGD
focusgroupdiscussion;QoLqualityoflife;GD
gooddeath;EoLC
end-of-lifecare.
© 2016 John Wiley & Sons Ltd16
S. Rainsford et al.
definition of rurality was reported. The research qual-ity assessment is reported in Table 1.
Objectives of the included articles
To explore or describe the ‘good death’ conceptthrough the perspectives of rural residents was theobjective of seven studies (Grant et al. 2003, Counts &Counts 2004, van der Geest 2004, Wilson et al. 2009a,
b, Viellette et al. 2010, Knight 2014, Biggs 2014). Sevenstudies (Felt et al. 2000, Arnaert et al. 2009, Grant et al.2011, Gysels et al. 2011, Beckstrand et al. 2012, 2015,Devik et al. 2013, Joarder et al. 2014) implied the‘good death’ by focusing on quality EoLC. The ‘gooddeath’ was not the primary objective in the remainingthree studies (Easom et al. 2006, Huy et al. 2007,Cruickshank et al. 2010) but was a point of commentby rural participants noted within the article text.
Methods of data collection
All eligible studies were qualitative in nature.Twenty-nine focus groups and 378 interviews wereconducted; 245 written surveys were completed; eightclinical encounters were detailed and six communitieswere observed. Sample sizes ranged from four(Knight 2014) to whole communities (Counts &Counts 2004, van der Geest 2004, Easom et al. 2006,Joarder et al. 2014). With the exception of one urban–rural study (Knight 2014), all the included studieswere classified as being solely rural in focus.
Participants
The exact number of participants is unknown as twoanthropological fieldwork studies included wholecommunities (Counts & Counts 2004, Easom et al.2006). Of the 751 identified rural participants, 84 wererural patients, 68 rural FCGs, 323 rural HCPs, 153rural residents, 83 rural community leaders and 40rural health administrators and policy makers. Of theidentified informants, 20% were patients or FCGs andof these 76% were African. HIV/AIDS (considered astigmatised disease) or cancer accounted for most ter-minal illnesses in Africa. All participants were adultsand when reported, their ages ranged from 18 to94 years. Participants over 80 years of age were inter-viewed in Canada (Biggs 2014), Bangladesh (Joarderet al. 2014) and Kenya (Grant et al. 2003). Of the 19patients in developed nations, 17 had cancer, onedementia and one cerebral vascular disease.
Concept
All the eligible articles reported on the ‘good death’from the perspectives or experiences of rural resi-dents. The term ‘good death’ was used explicitly ineight titles (PNG [1], Africa [2], Britain [1], Canada[4]) and an additional six abstracts (USA [2], Africa[2], Bangladesh [1], Vietnam [1]). For those studiesnot using the exact term, all implied the ‘good death’within the report or used an alternative term such asthe ‘ideal’ death (Beckstrand et al. 2012, 2015), ‘the
Table 2 Characteristics of included studies
Developed
(n = 10)
Developing
(n = 8)
Date of publication
2000–2004 1 3
2005–2009 3 2
2010–2015 6 3
Countries
Developed (n = 10)
North America 7 0
United Kingdom 2 0
Norway 1 0
Developing (n = 7)
Africa 0 5
Bangladesh 0 1
Vietnam 0 1
Papua New Guinea 0 1
Source of article
Journal 9 8
Dissertation 1 0
Source discipline
Palliative/hospice care 7 3
Social Science and Medicine 0 2
Other nursing 2 0
Population health 0 1
Anthropology 0 1
Cross-cultural gerontology 0 1
Theology 1 0
Study objective
included ‘Good death’
4 3
Type of article
Qualitative (n = 17)
Phenomenological 3 0
Ethnographic 2 1
Open ended written survey 2 0
Anthropological field work 0 3
Not specified 3 3
Systematic review 0 1
Methods*
Interviews 7 4
Focus groups 5 3
Observations/field work 0 4
Written surveys 2 0
Clinical observations 0 3
Terminology
Good death 7 4
Peaceful 0 2
Other 3 1
Bad death 0 1
*Some studies used more than one method.
© 2016 John Wiley & Sons Ltd 17
Rural residents’ perspectives on the rural ‘good death’
best scenario for one’s own death’ (Felt et al. 2000),an ‘honourable death’ (Huy et al. 2007), ‘dying withdignity’ (Grant et al. 2011), ‘dying peacefully’ (vander Geest 2004, Grant et al. 2011), dying ‘the properway’ (van der Geest 2004) and ‘facing death in abrave manner’ (Devik et al. 2013). One study focusedon the ‘bad death’ (Joarder et al. 2014). The ‘gooddeath’ referred to the death event, the dying process,the meaning of death and the after-death concept.
The dominant theme, from both developed anddeveloping countries was that a ‘good death’ is onethat is peaceful, that is free of pain and without suffer-ing. Other themes describing the ‘good death’ includeda ‘controlled’ death (control over symptoms, place ofdeath, decision-making, manner of death and toremain independent) (Felt et al. 2000, Counts & Counts2004, Dilger 2008, Wilson et al. 2009a,b, Viellette et al.2010, Knight 2014); a ‘timely’ death (Counts & Counts2004) that is a death coming ‘naturally and after a longand well-spent life’ (van der Geest 2004, p. 899) and‘hopefully in my sleep’ (Felt et al. 2000, p. 405) afterhaving had opportunity to say goodbye to family; a‘dignified’ death by maintaining identity, self-worth,integrity and control (Wilson et al. 2009a, Devik et al.2013); a ‘social’ death such as to die within the commu-nity with family present (Wilson et al. 2009a,b) and a‘noble’ death such as through enduring the situation(Grant et al. 2003, Devik et al. 2013). Two articles (vander Geest 2004, Knight 2014) acknowledged the diffi-culty of defining a ‘good death’ as it is dependent onindividual interpretations, perspectives and priorities.
Despite the challenges of definition, the reviewteam determined that rural residents identified fivedimensions they considered important for facilitatinga ‘good death’ – physical, emotional, social, spiritualand cultural dimensions.
Physical (pain and symptom management)Good pain and symptom control was the overridingfactor reported to ensure a ‘good death’. Pain reliefwas central to maintaining quality of life through thedying journey, not only just for the patient but alsothe FCG (Cruickshank et al. 2010).
In both developed and developing communities,physical care also includes human touch (Grant et al.2011, Biggs 2014), ‘touching [wounds] helps put asmile on people’s faces’ (Ugandan nurse; Grant et al.2011, p. 10). In contrast, western participants ‘fear atechnological death as opposed to a good death’ (Feltet al. 2000, p. 401), while in Ghana, ‘death was peace-ful and no medical battle was fought to keep himalive’ (van der Geest 2004, p. 902).
While acknowledging the importance of symptomcontrol, participants in two studies (Cruickshank et al.
2010, Grant et al. 2011) highlighted the need for a holisticapproach to the ‘good death’ including ‘emotional, spiri-tual, social and practical care’ (Grant et al. 2011, p. 1).
EmotionalLiving well while dying (Viellette et al. 2010, Deviket al. 2013) and maintaining dignity, respect, self-worth, autonomy and possibly also a sense ofhumour were all reported as essential for facilitatinga ‘good death’. In order to maintain autonomy, ‘thedying person needed to be lucid, to be able to think,and to have enough energy and mental alertness toshare their thoughts and feelings’ (Wilson et al.2009b, p. 315). The distress of not respecting patients’wishes was voiced by a number of rural emergencyroom nurses who indicated that ‘one obstacle to pro-viding [quality] EoLC is not knowing the patient’swishes regarding the continuation of treatments ortests because of [their] inability to communicate’(Beckstrand et al. 2012, p. 16). Research participantsfrom both developed and developing countries felt itwas important to know the truth about their illness(Felt et al. 2000) so they could be ‘at peace with theirown death’ (van der Geest 2004, p. 908), ‘stop chasing[expensive] false hopes’ (Grant et al. 2003, p. 161) andprepare for death; however, not all families want thepatient to know the truth (Grant et al. 2003). Fear andanxiety were cited as impediments to peace (Deviket al. 2013).
SocialMost rural participants identified the importance offamily and community. While a minority wanted to‘die alone, but only after having had closure withloved ones so that death would be peaceful’ (Feltet al. 2000, p. 405), most rural residents in all regionsand cultures said it was important to have familyand significant people present because ‘the together-ness of the family members makes you feel they loveyou and are not abandoning you’ (Grant et al. 2003,p. 163). African participants talked of the importantrole that family and community plays after death inensuring specific rituals are carried out. Acceptanceof the death by family members and ‘permission todie from loved ones [is] very important to a peacefuldeath’ (Felt et al. 2000, p. 405).
Ensuring there is no unfinished business or unre-solved personal conflicts and reconciliation is vitalacross rural/remote cultures:
Before one dies conflicts should be ended and enemies rec-onciled, debts should be paid and promises fulfilled.Someone who has been able to achieve this, is ready for hisfinal departure. He is . . . a peacemaker, a person who isrespected by others. (van der Geest 2004, p. 908)
S. Rainsford et al.
© 2016 John Wiley & Sons Ltd18
The place of death has significance with mostrural informants wishing to die ‘at home, which isthe epitome of peacefulness, surrounded by childrenand grandchildren’ (van der Geest 2004, p. 899) as‘the home or home community . . . [is] the only placewhere the dying person [can] be close to the manypeople who have meaning for them’ (Wilson et al.2009b, p. 316). In Ghana, dying away from home isconsidered ‘bad’ and disgraceful; however, partialrestoration can be achieved by ‘bringing the deadbody home’ (van der Geest 2004, p. 909). Some par-ticipants considered these social elements a low prior-ity and thus relatively unimportant (Knight 2014).
SpiritualA spiritual dimension was identified as the opportu-nity to say goodbye (Counts & Counts 2004) andexpress values and priorities; finding meaning to lifeand death which can ‘help to give a dying personhope’ (Biggs 2014) and gaining a sense of dying well(Wilson et al. 2009b). Two theological studies (Knight2014, Biggs 2014) placed the priority of spiritual andemotional peace (maintaining a sense of humour; senseof dignity; being at peace with God) ahead of physicaland social care with opportunity to say ‘goodbye’ tosignificant people a higher priority than dying athome. Participants from Africa (Grant et al. 2003, vander Geest 2004) and Canada (Biggs 2014) remarked onthe importance of religious rituals and how they ‘expe-rienced the support of spiritual care through spiritualpractices such as singing hymns, reading scripture andHoly Communion’ (Biggs 2014, p. 124).
One of the most important spiritual elements was the accep-tance of death . . . that could be reached through a processof dying well. For this state to occur, the dying personneeded to find meaning in their life and also meaning intheir daily experience. (Wilson et al. 2009b, p. 315)
CulturalThe influence of cultural beliefs and values was evi-dent in most articles with rural Kenyan participantsillustrating how ‘powerful cultural traditions [make]it difficult for social needs to be met’ (Grant et al.2003, p. 163). Religion was found to be a componentof cultural beliefs and values, with Christian religionsmost commonly identified (Grant et al. 2003, 2011,Counts & Counts 2004, van der Geest 2004, Knight2014, Biggs 2014). However, in some developingcountries, while Christianity has ‘greatly influenced’(van der Geest 2004) the concept of the ‘good’ and‘bad’ death, it has not replaced, but instead has beeninterpreted and applied to traditional beliefs (Grant
et al. 2003, 2011, Counts & Counts 2004, van derGeest 2004, Huy et al. 2007). In Vietnam (Huy et al.2007), Papua New Guinea (Counts & Counts 2004),Bangladesh (Joarder et al. 2014) and Ghana (van derGeest 2004), the ‘good’ and ‘bad’ death concepts werealigned with the circumstances surrounding the deathand afterlife, as these significantly impact on thehealth and well-being of the community, descendantsand ancestors. The ‘good death’ is also one that doesnot disrupt the life and health of the community (vander Geest 2004). Rural residents in developed coun-tries instead viewed the ‘good death’ in the light of abio-medical model, placing greater emphasis onautonomy, the process of dying and minimising anysense of struggle. The afterlife provides the dyingperson with a ‘hope of something more to come’(Biggs 2014). The culture of ‘not complaining’ wasreported in both developed (Devik et al. 2013) anddeveloping countries (Grant et al. 2003). ‘Povertyshaped how people died’ (Grant et al. 2011, p. 5) inrural sub-Saharan Africa.
Context
The overriding theme is that rural residents prefer todie at home, and when this is not possible then intheir rural community (Wilson et al. 2009a,b, Vielletteet al. 2010, Biggs 2014). ‘Some would prefer to haveless care and fewer services’ (Viellette et al. 2010, p.163) than leave their community.
Wilson et al. (2009a,b) reported that rural residentsrecognised unique challenges of providing EoLC inrural areas. These include distance (Grant et al. 2003,2011, Wilson et al. 2009a,b, Cruickshank et al. 2010),lack of services and personnel (Wilson et al. 2009a,b,Viellette et al. 2010, Devik et al. 2013), emotional andphysical isolation for independent HCPs (Arnaertet al. 2009), lack of access to specialist palliative care(Wilson et al. 2009a,b) and the distress of caring forclose friends and colleagues (Arnaert et al. 2009, Beck-strand et al. 2012, 2015). It was obvious there aremany advantages as well, such as the deep concernof community for its members (Wilson et al. 2009a,b).The benefits include personal attention (Arnaert et al.2009, Wilson et al. 2009a,b, Beckstrand et al. 2012,2015, Biggs 2014) and ‘the friendliness and closer-knitnature of a rural setting . . . the increased level of con-cern persons have for one another, and the resourcesavailable’ (Biggs 2014, p. 139). The rural environmentwas ‘identified as a much quieter and contemplativesetting’; however, ‘dying outside the rural communityis a common and highly problematic issue for ruralpeople’ (Wilson et al. 2009b, p. 317).
© 2016 John Wiley & Sons Ltd 19
Rural residents’ perspectives on the rural ‘good death’
Discussion
Principal findings
This scoping review identified 20 articles describingthe concept of the ‘good’ rural death from the per-spectives of a cross-section of rural residents includ-ing rural patients with a life-limiting illness, ruralfamily members/informal caregivers, rural HCP andthe wider rural community; and identified gaps inthe literature for areas for future studies to betterunderstand the concept of ‘good death’ for rural peo-ple. Both developed and developing countries wererepresented, with most data coming from NorthAmerica and sub-Saharan Africa.
Despite the challenges of rural definition, andnotwithstanding differences in rural locations, culturalperspectives, priorities and expectations, this scopingreview found similarities and differences in perspec-tives with those reported in urban studies. The fourelements (physical, spiritual, emotional and social)considered essential by the WHO (2015) to facilitate a‘good death’ were identified as pain/symptom con-trol, dignity, preparedness, autonomy and communityand are consistent with urban findings (Holdsworth2015, Raisio et al. 2015, Davies et al. 2016, Meier et al.2016). However, the context and priority placed oneach factor differed between included studies andfrom urban perspectives (Kirby et al. 2016).
While death is a normal end to life (WHO 2015),one could argue that many deaths are not entirely‘good’ due to the nature of the illness (Holdsworth2015), the age of the dying person (Counts & Counts2004, van der Geest 2004, Huy et al. 2007, Dilger2008) and individual persons’ perspective. The goalshould be to achieve the ‘best possible death giventhe circumstances’ (Holdsworth 2015) or a ‘goodenough’ death (McNamara 2004) to reflect a deaththat satisfies realistic expectations (Raisio et al. 2015).
Locality was also identified as a critical element,with deaths and dying ideally to be connected with therural/remote community. The literature increasinglyreports that place of death is one measure of a contem-porary ‘good death’ as supported by the majority ofarticles in this review; however as Biggs (2014) sug-gests, place is only one factor of importance and notnecessarily the main one for all people (Hoare et al.2015, Davies et al. 2016, Rainsford et al. 2016). If it isnot possible to die at home surrounded by family thenit is important to die within the rural community.
Knight (2014) suggests that the quality of dying issubjective and dependent on whose perspective one isusing. Family and friends are thought to perceive a‘good death’ has occurred when ‘timing, symptom
management and support come together successfully’(Adamson & Cruickshank 2013). Unfortunately due toheterogeneity within a small number of articles, it wasnot possible to compare the different perspectives ofrural informants or to compare with any confidencethe views of respondents in rich and poor countries;however, what was apparent is the different expecta-tions as in Kenya one is focused on the ‘basics of life’(Grant et al. 2003), whereas in Kansas, one fears ‘a tech-nological death’ (Felt et al. 2000). Good pain control ishighly ranked in EoLC (Byock et al. 2009); however,the difficulty of patients in developing countries access-ing basic drugs such as morphine was highlighted.
Wilson et al. (2009a,b) found that rural peopleidentify themselves as ‘rural’ with unique needs fromurbanites; and dying in a rural area was reported tohave elements that either enabled or obstructed a‘good death’. These elements included distances, iso-lation, limited services and personnel, strong sense ofcommunity and resilience. As such, the findings ofthis scoping review are diverse, rich and informative,but not uniform across the countries or studies.
Study meaning
While adding to rural end-of-life knowledge by iden-tifying and synthesising the current literature, thisscoping review highlights the paucity of informationrelating to the ‘good’ rural death especially from theperspectives of dying patients and their family mem-bers. Only 20% of the rural informants identified inthis review were patients or FCGs, with the majorityhaving a cancer or HIV/AIDS diagnosis, yet ‘toremain socially relevant, EoLC ideally must reflectthe needs of (all) the dying individuals . . . withindiverse cultural and geographic areas’ (Cottrell &Duggleby 2016, p. 26). It is not just the rural voicethat is as yet unheard and unrecognised as unique,but those with a non-malignant diagnosis as well asminority groups within rural communities. However,the findings of this review may be used to informfuture researchers and policy makers of the ‘quiet’rural voice and guide future planning and develop-ment of rural EoLC.
Limitations
This scoping review had limitations in the complete-ness of the review and the quality of identified studies.Studies not including the selection criteria wording inthe title, abstract or keywords may have been missed;however, electronic searching was augmented byhand-searching journals and reference lists. A poten-tial bias in data extraction, synthesis and analysis is
S. Rainsford et al.
© 2016 John Wiley & Sons Ltd20
also possible due to most authors being PalliativeMedicine Specialists and/or General Practitioners andall citizens of western nations. Currently, there is nobroad consensus on scoping review reporting guideli-nes, which potentially limits the report.
A more serious limitation is the lack of homogene-ity and consistency in the ‘rural’ definition. Thismeant that studies reported on a variety of rural loca-tions and rural populations. Extremely remote areaswere not studied, although some of the rural peoplesstudied may have been in extremely remote areaswith limited health services. Equally significant is thediversity in cultural perspectives. While this is astrength in that it presents a wide range of perspec-tives on the ‘good death’, it is also a limitation as itreduces the ability to examine the specific under-standing of the ‘good death’ within diverse cultures.
The objectives of some studies were not ‘gooddeath’ focused, a factor that increased the risk of ourmisinterpretation of the findings. Other limitationsinclude small rural sample size; inability to determineperspectives of rural informant subgroups; limitedgeneralisability; recruitment biases; a predominantlyChristian or traditional cultural viewpoint; and varia-tions in terminology.
Conclusion
Previous studies have found that rural residents con-sider themselves rural and as such have unique EoLCneeds and challenges. The existence of 20 papers indi-cates that rural perspectives on the ‘good death’ areimportant to consider. The current literature is hetero-geneous and thus insufficient to confidently gener-alise the perspectives of rural residents on the ‘gooddeath’ and what it means to die well and to avoiddying badly. Rural residents of both developed anddeveloping nations want a dignified death; however,this is largely subjective and thus open to interpreta-tion based on personal, cultural, social and religiousperspectives that could be highly individualised, fam-ily contextualised or rural community based. Clearly,given the importance of a ‘good death’, there is aneed for further studies, both qualitative and quanti-tative, to elicit rural person, especially dying patientsand family caregivers’ perspectives.
Author contribution
SR was responsible for conception; design; acquisition,analysis and interpretation of data; drafting, revisingand final manuscript. RDM, DW, NJG contributed toreview design. All co-authors contributed to analysisand interpretation of data; revising the article critically
for important intellectual content; and final approvalof the version to be published.
References
Adamson E. & Cruickshank S. (2013) A ‘good death’ athome: community nurses helping to make it possible. Bri-tish Journal of Community Nursing 18 (1), 40–42.
Aries P. (1974) Western Attitudes Towards Death: From theMiddle Ages to the Present. The John Hopkins UniversityPress, Baltimore.
Arksey H. & O’Malley L. (2005) Scoping studies: towards amethodological framework. International Journal of SocialResearch Methodology 8 (1), 19–32. Available at: http://eprints.whiterose.ac.uk/1618 (accessed on 12/2/2016).
Arnaert A., Seller R. & Wainwright M. (2009) Homecarenurses’ attitudes toward palliative care in a rural commu-nity in western Quebec. Journal of Hospice and PalliativeNursing 11 (4), 201–208.
Australian Government Dept of Health and Aging (2008)Australian report on the audit of health workforce in rural andregional Australia; 2.5.2 Trends in population aging. Com-monwealth of Australia.
Bakitas M.A., Elk R., Astin M. et al. (2015) Systematicreview of palliative care in the rural setting. Cancer Con-trol 22 (4), 450–463.
Beckstrand R.L., Giles V.C., Luthy K.E., Callister L.C. &Heaston S. (2012) The last frontier: rural emergencynurses’ perceptions of end-of-life care obstacles. Journal ofEmergency Nursing 38 (5), E15–E25.
Beckstrand R.L., Rohwer J., Luthy K.E., Macintosh J.L.B. &Rasmussen R.J. (2015) Rural emergency nurses’ end-of-lifecare obstacle experiences: stories from the last frontier.Journal of Emergency Nursing. Epub ahead of print.doi:10.1016/j.jen.2015.08.017.
Biggs D.D. (2014) “Comfort, comfort my people”: a qualita-tive study examining dying persons’ and their caregivers’experience of spiritual care as an aid to having a gooddeath in Grey and Bruce counties. Ontario DAI Section A.Humanities and Social Science 75, 11-A(E).
Byock I.R., Corbeil Y.J. & Goodrich M.E. (2009) Beyond polar-ization, public preferences suggest policy opportunities toaddress aging, dying, and family caregiving. American Jour-nal of Hospice and Palliative Care 26 (3), 200–208.
Colquhoun H.L., Levac D., O’Brien K.K. et al. (2014) Scop-ing reviews: time for clarity in definition, methods andreporting. Journal of Clinical Epidemiology 67, 1291–1294.
Cottrell L. & Duggleby W. (2016) The “good death”: anintegrative literature review. Palliative and Supportive Care.Epub ahead of print. doi:10.1017/S1478951515001285.
Counts D.A. & Counts D. (2004) The good, the bad, and theunresolved death in Kaliai. Social Science and Medicine 58,887897.
Cruickshank S., Adamson E., Logan J. & Brackenbridge K.(2010) Using syringe drivers in palliative care within arural, community setting: capturing the whole experience.International Journal of Palliative Nursing 16 (3), 126–132.
Daudt H.M.L., Van Mossel C. & Scott S.J. (2013) Enhancingthe scoping study methodology: a large, inter-professionalteam’s experience with Arksey and O’Malley’s frame-work. BMC Medical Research Methodology 13, 48.
Davies A., Todd J., Bailey F., Gregory A. & Waghorn M.(2016) Good concordance between patients and their non-
© 2016 John Wiley & Sons Ltd 21
Rural residents’ perspectives on the rural ‘good death’
professional carers about factors associated with a ‘gooddeath’ and other important end-of-life decisions. BMJ Sup-portive and Palliative Care. Epub ahead of print.doi:10.1136/bmjspcare-2015-001085.
Devik S.A., Enmarker I., Wiik G.B. & Hellzen O. (2013)Meanings of being old, living on one’s own and sufferingfrom incurable cancer in rural Norway. European Journalof Oncology Nursing 17 (6), 781–787.
Dilger H. (2008) “We are all going to die”: kinship, belong-ing, and the morality of HIV/AIDS-related illnesses anddeaths in rural Tanzania. Anthropological Quarterly 81 (1),207–232.
Downing J. & Jack B. (2012) End-of-life care in rural areas:what is different? Current Opinion in Supportive and Pallia-tive Care 6, 391–397.
Easom L.R., Galatas S. & Warda M. (2006) End-of-life care:and educational intervention for rural nurses in southeast-ern USA. International Journal of Palliative Nursing 12 (11),526–534.
Equator Network (2016) Preferred Reporting Items for SystematicReviews and Meta-Analysis extension for Scoping Reviews(PRISMA-ScR). Available at: http://www.equator-network.org/library/reporting-guidelines-under-development/#55;http://www.equator-network.org/wp-content/uploads/2009/02/Executive-summary_ScR_Dec-9.pdf (accessed on 27/2/2016).
Felt D.H., Early J.L. & Welk T.A. (2000) Attitudes, values,beliefs, and practices surrounding end-of-life care inselected Kansas communities. American Journal of Hospitaland Palliative Care 17 (6), 401–406.
van der Geest S. (2004) Dying peacefully: considering gooddeath and bad death in Kwahu-Tafo, Ghana. Social Scienceand Medicine 58, 899–911.
Gomes B. & Higginson I.J. (2008) Where people die (1974-2030): past trends, future projections and implications forcare. Palliative Medicine 22, 33–41.
Gomes B., Calanzani N., Gysels M., Hall S. & Higginson I.J.(2013) Heterogeneity and changes in preferences for dying athome: a systematic review. BMC Palliative Care 12 (7), 1–13.
Grant E., Murray S.A., Grant A. & Brown J. (2003) A gooddeath in rural Kenya? – Listening to Meru patients andtheir families talk about care needs at the end of life. Jour-nal of Palliative Care 19 (3), 159–167.
Grant L., Brown J., Leng M., Bettega N. & Murray S.A.(2011) Palliative care making a difference in ruralUganda, Kenya and Malawi: three rapid evaluation fieldstudies. BMC Palliative Care 10 (8), 1–10.
Gysels M., Pell C., Straus L. & Pool R. (2011) End of lifecare in sub-Saharan Africa: a systematic review of thequalitative literature. BMC Palliative Care 10 (6), 1–10.
Hoare S., Morris Z., Kelly M., Kuhn I. & Barclay S. (2015)Do patients want to die at home? A systematic review ofthe UK literature, focussed on missing preferences forplace of death. PLoS ONE 10 (11), e0142723.
Holdsworth L.M. (2015) Bereaved carers’ accounts of theend of life and the role of care providers in a ‘good death’:a qualitative study. Palliative Medicine 29 (9), 834–841.
Huy T.Q., Johansson A. & Long N.H. (2007) Reasons fornot reporting deaths: a qualitative study in rural Vietnam.World Health and Population 9 (1), 14–23.
JBI (2015) Joanna Briggs Institute Reviewers’ manual: method-ology for JBI scoping reviews. Joanna Briggs Institute.
Available at: www.joannabriggs.org (accessed on 12/2/2016).
Joarder T., Cooper A. & Zaman S. (2014) Meaning of death:an exploration of perception of elderly in a Bangladeshivillage. Journal of Cross Cultural Gerontology 29, 299–314.
Kellehear A. (2008) A Social History of Dying. CambridgeUniversity Press, New York.
Kirby S., Barlow V., Saurman E., Lyle D., Passey M. & Cur-row D. (2016) Are rural and remote patients, families andcaregivers needs in life-limiting illness different fromthose of urban dwellers? A narrative synthesis of the evi-dence. Australian Journal of Rural Health. Epub ahead ofprint. doi:10.1111/ajr.12312.
Knight D.A. (2014) What constitutes a good death? A clergyperspective. Rural Theology 12 (1), 2–16.
Levac D., Colquhoun H. & O’Brien K.K. (2010) Scopingstudies: advancing the methodology. ImplementationScience 5, 69. doi:10.1186/1748-5908-5-69.
McNamara B. (2004) Good enough death: autonomy andchoice in Australian palliative care. Social Science and Med-icine 58, 929–938.
Meier E., Gallegos J., Montross-Thomas L., Depp C., IrwinS. & Jeste D. (2016) Defining a good death (successfuldying): literature review and a call for research andpublic dialogue. American J Geriatric Psychiatry 24 (4),261–271.
Moher D., Liberati A., Tetzlaff J. et al. (2009) Preferredreporting items for systematic reviews and meta-analyses:the PRISMA statement. Annals of Internal Medicine 151 (4),264–270.
Peters M.D.J., Godfrey C.M., Khalil H., McInerney P., Par-ker D. & Soares C.B. (2015) Guidance for conducting sys-tematic scoping reviews. International Journal of EvidenceBased Healthcare. 13, 141–146.
Rainsford S., MacLeod R. & Glasgow N. (2016) Place ofdeath in rural palliative care: a systematic review. Pallia-tive Medicine 30 (8), 745–763.
Raisio H., Vartiainen P. & Jekunen A. (2015) Defining agood death: a deliberative democratic view. Journal of Pal-liative Care 31 (3), 158–165.
Spice R., Paul L.R. & Biondo P.D. (2012) Development of arural palliative care program in the Calgary zone ofAlberta Health services. Journal of Pain and Symptom Man-agement 43 (5), 911–924.
Tricco A.C., Lille E., Zarin W. et al. (2016) A scoping reviewon the conduct and reporting of scoping reviews. BMCMedical Research Methodology 16, 15.
Viellette A., Fillion L., Wilson D.M., Thomas R. & DumontS. (2010) La belle mort en milieu rural: a report of anethnographic study of the good death for Quebec ruralfrancophones. Journal of Palliative Care 26 (3), 159–166.
WHO (2015) World Health Organization definition of palliativecare. Available at: http://www.who.int/cancer/pallia-tive/definition/en/ (accessed on 28/11/2015).
Wilson D.M., Fillion L., Thomas R., Justice C., BhardwajP.P. & Veillette A. (2009a) The “good” rural death: areport of an ethnographic study in Alberta, Canada. Jour-nal of Palliative Care 25 (1), 21–29.
Wilson D.M., Fillion L., Thomas R., Justice C., Veillette A. &Bhardwaj P.P. (2009b) Planning and providing for a gooddeath using rural French-Canadian and English-Canadianinsights. Reviews in the Neurosciences 20, 313–319.
© 2016 John Wiley & Sons Ltd22
S. Rainsford et al.
https://doi.org/10.1177/0269216316685234
Palliative Medicine2017, Vol. 31(10) 895 –912© The Author(s) 2017Reprints and permissions: sagepub.co.uk/journalsPermissions.navDOI: 10.1177/0269216316685234journals.sagepub.com/home/pmj
Rural end-of-life care from the experiences and perspectives of patients and family caregivers: A systematic literature review
Suzanne Rainsford1, Roderick D MacLeod2,3, Nicholas J Glasgow1, Christine B Phillips1, Robert B Wiles4 and Donna M Wilson5
AbstractBackground: End-of-life care must be relevant to the dying person and their family caregiver regardless of where they live. Rural areas are distinct and need special consideration. Gaining end-of-life care experiences and perspectives of rural patients and their family caregivers is needed to ensure optimal rural care.Aims: To describe end-of-life care experiences and perspectives of rural patients and their family caregivers, to identify facilitators and barriers to receiving end-of-life care in rural/remote settings and to describe the influence of rural place and culture on end-of-life care experiences.Design: A systematic literature review utilising the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines.Data sources: Four databases (PubMed, CINAHL, Scopus and Web of Science) were searched in January 2016, using a date filter of January 2006 through January 2016; handsearching of included article references and six relevant journals; one author contacted; pre-defined search terms and inclusion criteria; and quality assessment by at least two authors.Results: A total of 27 articles (22 rural/remote studies) from developed and developing countries were included, reporting rural end-of-life care experiences and perspectives of patients and family caregivers. Greatest needs were informational (developed countries) and medications (developing countries). Influence of rural location included distances, inaccessibility to end-of-life care services, strong community support and importance of home and ‘country’.Conclusion: Articulation of the rural voice is increasing; however, there still remain limited published rural studies reporting on patient and family caregivers’ experiences and perspectives on rural end-of-life care. Further research is encouraged, especially through national and international collaborative work.
KeywordsPalliative care, end-of-life care, rural, family, patient, caregiver, systematic review
1 Medical School, Australian National University, Canberra, ACT, Australia
2HammondCare, Sydney, NSW, Australia3Palliative Medicine, University of Sydney, Sydney, NSW, Australia4 Rural Clinical School, Australian National University, Cooma, NSW, Australia
5Faculty of Nursing, University of Alberta, Edmonton, AB, Canada
Corresponding author:Suzanne Rainsford, Medical School, Australian National University, C/- PO Box 159, Cooma, 2630, NSW, Australia. Email: suzanne.rainsford@anu.edu.au
685234 PMJ0010.1177/0269216316685234Palliative MedicineRainsford et al.review-article2016
Review Article
What is already known about the topic?
•• Quality end-of-life care is the right of every person regardless of where they live.•• Rural areas are distinct and as such need to have special consideration.•• Gaps exist in rural end-of-life care research including studies reporting the experiences and perspectives of rural
patients and their family caregivers.
896 Palliative Medicine 31(10)
Introduction
Quality end-of-life care is considered the right of every dying person regardless of where they live.1 Quality end-of-life care is patient-centred care where patients and their primary carer, often a family member, are encouraged to participate, if not direct, care decisions as ‘to remain socially relevant, end-of-life care ideally must reflect the needs of the dying individual …. within diverse cultural and geographic areas (p. 26)’.2 Rural areas are distinct and as such need to have special consideration. It is imperative to take full account of the rural experiences and perspec-tives of those persons and families receiving care to ensure optimal rural end-of-life care.3
Wilson et al.4,5 suggest rural people define themselves as ‘rural’ and as different from urbanites; however, gaining the rural voice is challenging. One significant obstacle in finding a common ‘rural’ voice is the lack of an interna-tionally agreed rurality index, meaning that results across studies and across countries are not necessarily compara-ble. Hart et al.6 report ‘rural’ to be a ‘multifaceted concept (p. 1149)’6 dependent on context. As such, there are incon-sistencies in definitions of ‘rural’ based on population size, density or demographics; or distance from urban centres and services; or as a specific ‘culture’.6 An additional research challenge is the difficulty of recruiting terminally ill patients and their family caregivers to research7 which is magnified by the often small cohort of such patients within small rural communities.
Previous rural palliative care reviews3,8 report that rural research is focused on programme planning, integration and evaluation; education; finances; and needs assess-ments,9 with professional providers or administrators as the prime informants. In 2009, Robinson et al.8 published a systematic review (studies published from 1996 through 2007) identifying major gaps in the literature including a lack of studies describing end-of-life care through the experiences and perspectives of rural patients and family caregivers and the influence of rural culture on these expe-riences. Future researchers were encouraged to seek these perspectives in order to provide ‘strong evidence to inform
palliative care policy and service development in rural set-tings (p. 253)’.8
The objectives of our systematic review were to search the literature since January 2006 and to (1) explore the end-of-life care experiences and perspectives of rural patients and their family caregivers, (2) identify facilitators and bar-riers to receiving end-of-life care in rural/remote settings and (3) describe the influence of rural place and culture on end-of-life care experiences. This information will add to the general knowledge on rural/remote end-of-life care and may further assist rural policymakers and healthcare professionals (HCPs) in ensuring that rural palliative care services are relevant to those receiving such care.
Method
This systematic review was undertaken utilising the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines.10
Literature search
In January 2016, four electronic databases (PubMed, CINAHL, Scopus and Web of Science) were searched using the following keywords and Medical Subject Heading (MeSH) terms: (‘palliative care’ OR hospice OR ‘terminal care’ OR end-of-life OR ‘end-of-life care’) AND (Rural OR Remote) AND (patient OR family OR carer OR caregiver) AND (perspective OR experience OR opinion OR view). A date filter of January 2006 through January 2016 was applied with this including an overlap with the 2009 systematic review8 to ensure that previously submit-ted but unpublished studies were included. Two articles (2006 and 200711) were identified and considered; how-ever, as the 2006 article was included in the previous review, it was excluded.
The reference lists of all included studies were scanned for additional articles. Recent issues (July 2014–May 2016) of six relevant journals (Palliative Medicine, Journal of
What this paper adds?
•• Rural patients and family caregivers experience unique challenges and benefits of rural living.•• The greatest needs of patients and family caregivers were informational (developed countries) and medications (devel-
oping countries).•• Influences of rural location on end-of-life care included distances, inaccessibility to end-of-life care services, strong com-
munity support and importance of home and ‘country’.•• Articulation of the rural voice is increasing, especially in North America.
Implications for practice, theory or policy
•• Further research is encouraged to gain the experiences and perspectives of rural patients and family caregivers, espe-cially through national and international collaborative work.
Rainsford et al. 897
Palliative Medicine, Journal of Pain and Symptom Management, Palliative and Supportive Care, Australian Journal of Rural Health and Journal of Rural Health) were handsearched by the first author (S.R.).
Selection criteria
Studies were assessed against predetermined inclusion criteria and included rural or remote residents (no stand-ardised definition of rurality was used); participants were receiving palliative or end-of-life care for malignant or non-malignant illnesses; data pertained to patient or fam-ily caregivers’ experiences or perspectives on end-of-life care collected from interviews or surveys. Both qualita-tive and quantitative studies were included. No age filter was applied. Only published international English lan-guage, peer-reviewed research articles were considered. Articles that included urban and rural data were included providing the rural data were clearly identifiable. One author11 of a retained study was contacted to verify par-ticipants were rural.
Rejected studies were either clearly irrelevant or those that addressed the topic in general but failed in one or more of the inclusion criteria. Ethics approval was not required for this review of published literature.
Screening of papers
The search identified 450 potential papers. After dupli-cates were removed, 220 titles were screened indepen-dently by S.R. and second author (R.D.M.) against the inclusion criteria. This identified 105 papers for considera-tion. At least two authors screened each abstract, and when necessary, full texts. Differences were discussed via email and resolved by consensus.
Appraisal and data extraction
Quality assessment of each retained article was made by S.R. with verification shared independently among the five co-authors. Quality was determined on aspects rele-vant to rural patient/family caregiver perspectives on end-of-life care and was not necessarily an assessment of the study per se, which resulted in some high-quality studies receiving a lower score. All studies were rated to be of low, medium or high quality based on a simple scoring system described by Gomes et al.12 and modified by S.R. to account for the rural patient/family caregiver focus. As such, two additional items were included as follows: (1) clarity of rural definition and (2) validity of informant (first hand/prospective = 2, retrospective family car-egiver = 1). Only studies rated high or medium were included in this literature review.13
Data from each retained article were extracted and tabled by S.R., according to pre-defined categories (see
Table 1). Critical review of full texts against data extrac-tion was shared and undertaken independently among the five co-authors.
Data synthesis
Study findings were coded into four categories: (1) patient perspectives, (2) family caregivers’ perspectives, (3) facil-itators and barriers and (4) influence of rural place and cul-ture. Each category was analysed thematically and reported descriptively. Due to heterogeneity, a meta-analysis was not possible. Differences in quality assessment, data extraction and synthesis were discussed by email and resolved by consensus.
Results
Overview of findings
Of the initial 450 articles identified and screened, a total of 27 articles11,14–39 reporting on 22 separate studies met our inclusion criteria (see Figure 1 for PRISMA flowchart40). In total, 19 studies (86%) were conducted in developed coun-tries: Canada (7),14,15,22–24,30–32,35,36 United States (5),17,26,33,34,37 Australia (2),18,27 Scotland (2),16,29 Norway (2),19–21 and England (1);11 and 3 in sub-Saharan Africa including Malawi (1),28 Cameroon (1)38 and one multi-nation study (Malawi, Kenya, Uganda).25 Three studies focused on indigenous pop-ulations (Australia,18 Canada30 and United States33). Rural definitions, locations and degree of isolation between studies were heterogeneous; four studies18,25,29,31,37 included remote locations. Studies were published between 2007 and 2015; no published articles were identified in 2016. In all, 18 stud-ies were qualitative, 3 quantitative,14,15,17,24 and 1 mixed-methods.31,32 Qualitative methods were described as ethnographic (4),18,25,33,35,36 phenomenological (4),19–21,30,37 grounded theory (1),11 and non-specified (9). Studies col-lected data from interviews (face-to-face (17), phone (1)22,23); surveys (written (3),17,24,31,32 phone (1)14,15); focus groups (1),34 and analysis of personal daily journals (1).39 Sample size ranged from 519,20 to 672.17
Participants included those with cancer and non-cancer diagnoses; active and bereaved family caregivers. A cross section of participant ages was found, with one paediatric study38 and eight studies where all patients (participants or deceased) were elderly (⩾60 years).11,16,17,19–23,26,33 Patient and family caregiver characteristics are reported in Tables 2 and 3.
Three studies explored the experiences and perspec-tives of patients only;16,19–21 eight of family caregivers only;14,15,17,24,26,27,30,37,39 eight of both patients and family caregivers;11,18,22,23,28,29,31,32,38 and three were community-focused studies including family caregivers.25,34–36 One study compared urban and rural perspectives,14,15 one the congruence between patient and family caregiver views,38
898 Palliative Medicine 31(10)
Tab
le 1
. Su
mm
ary
of in
clud
ed s
tudi
es.
Purp
ose
Info
rman
ts/r
ecru
itmen
tR
ural
def
initi
onM
etho
dolo
gyFi
ndin
gsLi
mita
tions
/bia
sQ
ualit
y
Braz
il et
al.1
4,15
Can
ada
To
com
pare
exp
erie
nces
an
d su
ppor
t ne
eds
of
urba
n an
d ru
ral F
CG
, to
dete
rmin
e im
port
ance
of
and
degr
ee o
f unm
et n
eeds
Mix
ed g
eogr
aphi
cal s
tudy
140
FCG
s (in
clud
ed 7
0 ru
ral F
CG
s: 4
4 ac
tive +
26
bere
aved
)R
ecru
itmen
t: id
entif
ied
by
sing
le P
C s
ervi
ceR
R: 2
9%
Rur
al =
area
out
side
ur
ban
(com
mun
ities
⩾
30,0
00 r
esid
ents
); ru
ral a
nd is
olat
ed
com
mun
ities
in
Nor
thea
ster
n O
ntar
io
Cro
ss-s
ectio
nal t
elep
hone
sur
vey;
de
mog
raph
ics;
mod
ified
val
idat
ed
inst
rum
ents
: CBS
-EoL
, MSP
SS,
ECO
G, c
atal
ogue
ser
vice
s av
aila
ble,
25
-item
car
er s
uppo
rt n
eed
inst
rum
ent +
5 qu
estio
ns r
elat
ing
to
serv
ice
deliv
ery;
ana
lysi
s–de
scri
ptiv
e an
d in
fere
ntia
l sta
tistic
s; e
valu
atio
n by
loca
tion
(urb
an v
s ru
ral)
Mos
t im
port
ant
need
s fo
r ur
ban
(U) +
rur
al
(R)
wer
e in
form
atio
nal;
rura
l FC
G: g
reat
er
unm
et t
angi
ble
need
s (z
= −
2.59
, p =
0.0
1);
emot
iona
l sup
port
U =
R; c
are
perc
eive
d as
la
rgel
y ac
cess
ible
to
patie
nts’
cha
ngin
g ne
eds
for
U +
R r
espo
nden
ts (
mea
n ac
cess
ibili
ty
scor
e: 1
.7, S
D =
0.7
and
1.9
, SD
= 0
.09,
re
spec
tivel
y, t8
0 =
1.3
1, p
= 0
.20,
95%
CI:
0.12
–0.5
8, d
= 0
.29)
Low
res
pons
e ra
te,
recr
uitm
ent
thro
ugh
palli
ativ
e ca
re s
ervi
ce, r
ural
re
side
ncy
dive
rse,
eva
luat
ion
of s
ingl
e se
rvic
e, b
ias
as
over
-bur
dene
d ca
rers
may
no
t ha
ve p
artic
ipat
ed, s
urve
y di
rect
ed t
owar
ds e
xist
ing
serv
ice
deliv
ery
Hig
h
Dal
e an
d Jo
hnst
on16
Scot
land
To
expl
ore
and
inte
rpre
t th
e co
ncer
ns o
f pat
ient
s w
ith in
oper
able
lung
ca
ncer
, to
info
rm s
peci
alis
t pa
lliat
ive
care
nur
sing
pr
actic
eR
ural
stu
dy
Six
patie
nts
Rec
ruitm
ent:
purp
osiv
e sa
mpl
e; id
entif
ied
by lu
ng
canc
er n
urse
spe
cial
ist
Dis
ease
: lun
g ca
ncer
Rur
al c
omm
unity
in
Scot
land
Qua
litat
ive,
inte
rpre
tativ
e co
nstr
uctiv
ist
appr
oach
, sem
i-st
ruct
ured
exp
lora
tory
ITV
s,
them
atic
ana
lysi
sIn
terv
iew
er: r
esea
rche
r nu
rse,
un
know
n to
par
ticip
ants
The
mes
:1.
Ste
adfa
stly
livi
ng li
fe; m
akin
g m
ost
of t
ime
avai
labl
e, m
aint
aini
ng n
orm
ality
, tak
e lif
e as
it
com
es, n
ot g
ivin
g up
, not
bec
omin
g a
burd
en
on fa
mily
, mai
ntai
ning
iden
tity
and
valu
e; o
ne
part
icip
ant
indi
cate
d no
oth
er c
hoic
e bu
t ca
rry
on a
s no
rmal
2. F
amily
sup
port
and
sep
arat
ion:
sup
port
of
and
con
cern
for
fam
ily, p
lann
ing
so
no u
nfin
ishe
d bu
sine
ss, a
ntic
ipat
ed lo
ss,
prep
arat
ion
for
sepa
ratio
n3.
Tru
st in
pro
fess
iona
ls: s
ome
wan
ted
info
rmat
ion
to m
ake
deci
sion
s, o
ther
s co
nten
t to
follo
w a
dvic
e w
ithou
t qu
estio
ning
, kn
owin
g w
ho t
o co
ntac
t fo
r su
ppor
t
Smal
l sam
ple
size
, sin
gle
serv
ice
recr
uitm
ent
Hig
h
Dar
er e
t al
.17
USA
To
asse
ss q
ualit
y of
the
dy
ing
proc
ess,
iden
tify
oppo
rtun
ities
for
impr
ovem
ent
and
supp
ort
stra
tegi
es fo
r ta
rget
ing
at-
risk
fam
ilies
Rur
al s
tudy
672
NO
K (
bere
aved
)R
ecru
itmen
t: id
entif
ied
from
sin
gle
inte
grat
ed
rura
l hea
lth d
ata,
mai
led
surv
ey, o
pt-o
utD
iagn
osis
: All;
ESR
D,
CH
F, C
OPD
ove
rsam
pled
RR
: 25%
Cen
tral
and
NE
rura
l Pe
nnsy
lvan
iaW
ritt
en s
urve
y co
ntai
ned
QoD
D
inst
rum
ent,
GA
D-7
, PH
Q-8
, m
odifi
ed T
oolk
it of
Inst
rum
ents
to
mea
sure
EoL
CQ
uant
itativ
e an
alys
is–d
escr
iptiv
e,
com
pari
sons
(ch
i squ
are,
line
ar
regr
essi
on, p
val
ues)
In g
ener
al, h
igh
satis
fact
ion
with
car
e te
ams
and
prov
ider
s, lo
wer
QoD
D e
xper
ienc
e as
soci
ated
w
ith d
ying
in h
ospi
tal,
rece
ipt
of c
onfli
ctin
g in
form
atio
n, p
oor
co-o
rdin
atio
n be
twee
n H
CPs
and
con
fusi
on a
roun
d w
hich
doc
tor
is in
cha
rge,
NO
K d
epre
ssio
n an
d an
xiet
y,
prol
onge
d ill
ness
Poor
res
pons
e ra
te,
sing
le in
tegr
ated
hea
lth
syst
em, l
imite
d di
vers
ity o
f pa
rtic
ipat
ing
NO
K, u
p to
12
mon
ths
bere
avem
ent
(pot
entia
l rec
all b
ias)
, pr
edom
inan
tly W
hite
fe
mal
es ⩾
60 ye
ars,
NO
K n
ot
nece
ssar
ily r
ural
Med
ium
Dem
bins
ky18
Aus
tral
iaT
o ex
plor
e pe
rcep
tions
an
d us
e of
pal
liativ
e ca
re
serv
ices
by
exam
inin
g liv
ed
expe
rien
ces
of w
omen
with
br
east
can
cer
Rur
al/r
emot
e st
udy
10 p
atie
nts;
4 F
CG
, 6
non-
care
r fa
mily
m
embe
rs, 5
HC
PsR
ecru
itmen
t: R
egio
nal
Med
ical
Ser
vice
s an
d In
dige
nous
Wom
en’s
Su
ppor
t N
etw
ork
Dia
gnos
is: b
reas
t ca
ncer
Yam
atji
coun
try,
m
id-w
est
regi
on o
f W
este
rn A
ustr
alia
; sc
arce
ly p
opul
ated
with
fe
w s
mal
l tow
ns
Med
ical
ant
hrop
olog
y, in
form
al IT
Vs,
gr
ound
ed t
heor
y an
d et
hnog
raph
ic
imm
ersi
on, c
oded
, the
mat
ic a
naly
sis
Inte
rvie
wer
: non
-tre
atin
g re
sear
cher
Prep
arat
ion
– to
gat
her
fam
ily, d
istr
ibut
e po
sses
sion
s, s
et a
ffair
s in
ord
er, s
ay g
oodb
yeT
o di
e in
cou
ntry
– if
one
die
s ou
tsid
e th
eir
coun
try,
the
spi
rit
is d
oom
ed t
o ro
am e
tern
ally
, pr
even
ting
com
plet
ion
of li
fe–d
eath
–life
co
ntin
uum
Spec
ific
cere
mon
ies
and
prac
tices
sur
roun
ding
de
ath,
for
exam
ple,
ritu
al c
lean
sing
of p
hysi
cal
deat
h pl
ace
inhi
bits
upt
ake
of in
patie
nt c
are
Opp
ortu
nity
for
fam
ily t
o re
mai
n cl
ose
to t
he
dyin
g m
embe
r is
impo
rtan
t fo
r th
e su
rviv
ing
fam
ilyH
ospi
tal e
nvir
onm
ent
not
cultu
rally
app
ropr
iate
Hom
e-ba
sed
PC li
mite
d du
e to
dis
tanc
es,
mob
ile s
ervi
ce v
isit
1–2
times
/mon
ths
Smal
l sam
ple
size
, spe
cific
cu
ltura
l iss
ues,
onl
y fe
mal
e pa
rtic
ipan
ts w
ith b
reas
t ca
ncer
, rec
ruite
d th
roug
h si
ngle
hea
lth s
ervi
ce,
snow
balli
ng o
f par
ticip
ants
Hig
h
Rainsford et al. 899
Purp
ose
Info
rman
ts/r
ecru
itmen
tR
ural
def
initi
onM
etho
dolo
gyFi
ndin
gsLi
mita
tions
/bia
sQ
ualit
y
Dev
ik e
t al
.19
Wiik
et
al.20
Nor
way
To
expl
ore
and
unde
rsta
nd
lived
exp
erie
nce
of o
lder
ca
ncer
pat
ient
s liv
ing
alon
eR
ural
stu
dy
5 ol
der
patie
nts,
livi
ng
alon
e; (
3 w
omen
, 2 m
en,
aged
71-
79);
rece
ivin
g ou
tpat
ient
and
life
-pr
olon
ging
che
mot
hera
pyR
ecru
itmen
t: id
entif
ied
thro
ugh
onco
logy
po
lycl
inic
; ini
tial c
onta
ct
by n
urse
Dia
gnos
is: i
ncur
able
ca
ncer
RR
: 50%
Com
mun
ities
(m
ount
ain
and
coas
tal),
lo
w p
opul
atio
n de
nsity
, po
pula
tion:
920
–777
5
Qua
litat
ive;
phe
nom
enol
ogic
al
herm
eneu
tical
app
roac
h; n
arra
tive
ITV
sIn
terv
iew
er: N
on t
reat
ing
rese
arch
er
Endu
ring
by
keep
ing
hope
aliv
e, b
ecom
ing
awar
e th
at y
ou a
re o
n yo
ur o
wn,
livi
ng u
p to
ex
pect
atio
ns o
f bei
ng a
goo
d pa
tient
and
bei
ng
at r
isk
of lo
sing
one
’s id
entit
y an
d va
lue
Rur
al d
ista
nce
requ
ires
com
mut
ing
(str
essf
ul
com
mut
ing,
bei
ng e
xhau
sted
)
Smal
l sam
ple,
rec
ruitm
ent
bias
(po
lycl
inic
), ca
ncer
Hig
h
Dev
ik e
t al
.21
Nor
way
To
illum
inat
e an
d in
terp
ret
the
mea
ning
of r
ecei
ving
ho
me
nurs
ing
care
whe
n be
ing
old
and
livin
g w
ith
adva
nced
can
cer
Rur
al s
tudy
9 ol
der
patie
nts
(5
wom
en, 4
men
); 71
–92
year
sId
entif
ied
by o
ncol
ogy
and
canc
er c
are
nurs
esD
iagn
osis
: can
cer
A t
erri
tori
al a
rea
with
sc
atte
red
sett
lem
ents
, so
me
dist
ance
from
a
dens
ely
popu
late
d ar
ea a
nd w
here
na
tura
l res
ourc
es a
re
the
basi
s fo
r in
dust
ry
and
outd
oor
activ
ity;
part
icip
ants
res
ided
w
here
pop
ulat
ion:
26
8–50
89 (
mea
n 5
inha
bita
nts/
km2 )
Qua
litat
ive,
phe
nom
enol
ogic
al
herm
eneu
tical
app
roac
h, n
arra
tive
ITV
sIn
terv
iew
er: r
esea
rche
r
Rec
eivi
ng H
NC
mea
ns t
o co
ntin
ue a
fam
iliar
an
d m
eani
ngfu
l life
Thr
ee t
hem
es: b
eing
con
tent
with
wha
t on
e ge
ts (
a w
ish
to a
ge in
pla
ce),
falli
ng in
to p
lace
an
d lo
sing
one
’s p
lace
Stro
ng p
lace
att
achm
ent
(phy
sica
l, so
cial
, au
tobi
ogra
phic
al);
sugg
ests
rur
al a
dvan
tage
ous
heal
thca
re e
nvir
onm
ent
(com
fort
, sec
urity
, id
entit
y)
Smal
l sam
ple,
can
cer,
re
crui
tmen
t by
hom
e ca
re
nurs
esSt
reng
th: r
esea
rche
r re
flexi
vity
ack
now
ledg
ed
Hig
h
Dug
gleb
y et
al.2
2,23
Can
ada
To
expl
ore
the
cont
ext
and
expe
rien
ce o
f sig
nific
ant
tran
sitio
ns e
xper
ienc
ed
by o
lder
rur
al p
erso
ns
rece
ivin
g pa
lliat
ive
hom
e ca
re a
nd t
heir
fam
ilies
and
to
dev
elop
a s
ubst
antiv
e th
eory
of t
rans
ition
s in
thi
s po
pula
tion
Rur
al s
tudy
6 ol
der
patie
nts;
10
bere
aved
FC
Gs
Rec
ruitm
ent:
purp
osiv
e sa
mpl
e, id
entif
ied
by P
C
co-o
rdin
ator
Dia
gnos
is: a
dvan
ced
canc
er
3 ru
ral h
ealth
reg
ions
in
a w
este
rn C
anad
ian
prov
ince
cla
ssifi
ed a
s on
e of
the
mos
t ‘r
ural
’ C
anad
ian
prov
ince
s
Gro
unde
d th
eory
, ope
n-en
ded
indi
vidu
al a
udio
-tap
ed IT
Vs
(pat
ient
s –
face
-to-
face
; FC
G –
tel
epho
ne)
Inte
rvie
wer
: res
earc
her
‘Nav
igat
ing
Unk
now
n W
ater
s’ d
escr
ibes
tr
ansi
tions
face
d by
old
er r
ural
pal
liativ
e pa
tient
s an
d th
eir
fam
ilies
due
to
rura
l iso
latio
n,
lack
of i
nfor
mat
ion,
lim
ited
acce
ssib
ility
to
serv
ices
, ind
ivid
ual v
alue
s an
d co
mm
unity
co
nnec
tedn
ess
FCG
s ex
peri
ence
d m
ultip
le s
igni
fican
t tr
ansi
tions
incl
udin
g si
gnifi
cant
cha
nges
in t
heir
ow
n ph
ysic
al a
nd m
enta
l hea
lth
Smal
l sam
ple,
cro
ss-s
ectio
nal
surv
ey s
o po
int
in t
ime
data
, va
riab
le g
eogr
aphi
c lo
catio
n,
canc
er o
nly
Hig
h
Dug
gleb
y et
al.2
4
Can
ada
To
dete
rmin
e fa
ctor
s in
fluen
cing
the
hop
e of
ru
ral w
omen
car
egiv
ers
Rur
al s
tudy
122
FCG
(ac
tive)
Rec
ruitm
ent:
surv
ey
and
invi
tatio
n m
aile
d to
can
cer
patie
nts
(to
forw
ard
to p
rim
ary
fem
ale
care
r) Id
entif
ied
by
canc
er r
egis
try
with
60%
no
t re
ceiv
ing
serv
ices
Dia
gnos
is: c
ance
rR
R: 1
5.6%
Hom
e ad
dres
s w
ith
a ru
ral p
ost
code
, Sa
skat
chew
an ⩽
999
9 re
side
nts,
Alb
erta
<
55,0
00
A c
ross
-sec
tiona
l pro
spec
tive
corr
elat
iona
l des
ign,
wri
tten
sur
vey,
qu
antit
ativ
e an
alys
is (
SPSS
)
Men
tal h
ealth
, wel
l-bei
ng a
nd s
elf-e
ffica
cy w
ere
pred
icto
rs o
f hop
e; p
hysi
cal h
ealth
was
not
si
gnifi
cant
Gui
lt le
ssen
s ho
pe a
nd in
crea
ses
burd
en, d
epre
ssio
n an
d fa
mily
dis
tres
sH
ighe
r ho
pe s
core
s as
soci
ated
with
hig
her
men
tal h
ealth
sco
res,
low
er p
erce
ptio
ns o
f lo
ss a
nd g
rief
sco
res
and
high
er s
core
s in
the
ir
abili
ty t
o de
al w
ith d
iffic
ult
situ
atio
nsR
ural
wom
en c
areg
iver
s ha
d lo
w p
hysi
cal
and
men
tal h
ealth
sco
res
com
pare
d to
US
popu
latio
n no
rms
(<25
th p
erce
ntile
)
No
mea
sure
of h
ope
over
tim
e, s
mal
l sam
ple,
sm
all
resp
onse
rat
e, b
ias
in
recr
uitm
ent
– se
lf-ac
cept
ed
invi
tatio
n, c
ance
r
Med
ium
Tab
le 1
. (C
ontin
ued)
(Con
tinue
d)
900 Palliative Medicine 31(10)
Purp
ose
Info
rman
ts/r
ecru
itmen
tR
ural
def
initi
onM
etho
dolo
gyFi
ndin
gsLi
mita
tions
/bia
sQ
ualit
y
Gra
nt e
t al
.25
Uga
nda,
K
enya
, Mal
awi
To
desc
ribe
pat
ient
, fam
ily
(and
loca
l com
mun
ity)
pers
pect
ives
on
the
impa
ct
of t
hree
com
mun
ity-b
ased
PC
inte
rven
tions
in s
ub-
Saha
ran
Afr
ica
Rur
al s
tudy
33 p
atie
nts;
27
FCG
s (1
7 ac
tive,
10
bere
aved
); al
so
29 le
ader
s + 6
1 st
aff
Rec
ruitm
ent:
iden
tifie
d by
PC
tea
mD
iagn
osis
: adv
ance
d ill
ness
Uga
nda:
mob
ile P
C
serv
ice
120
km fr
om
Kam
pala
incl
udes
fish
ing
com
mun
ities
on
Lake
V
icto
ria
Ken
ya: r
ural
, 240
km
fr
om N
airo
biM
alaw
i: pe
ri-u
rban
di
stri
cts
of B
lant
yre
Rap
id a
sses
smen
t ev
alua
tion,
qu
alita
tive,
pho
togr
aphi
c et
hnog
raph
y, r
apid
eva
luat
ion
field
st
udie
s, IT
Vs
with
key
info
rman
ts,
dire
ct o
bser
vatio
ns o
f clin
ical
en
coun
ters
, (re
view
of l
ocal
PC
in
form
atio
n), d
ata
tria
ngul
atio
n,
them
atic
ana
lysi
sIn
terv
iew
er: r
esea
rche
r, in
terp
rete
r,
PC w
orke
r pr
esen
t
Patie
nts
valu
ed b
eing
tre
ated
with
res
pect
, m
aint
aini
ng d
igni
ty, i
nteg
ratio
n of
sym
ptom
co
ntro
l with
pra
ctic
al, e
mot
iona
l, fin
anci
al
and
spir
itual
car
e as
sist
ed p
atie
nts
to r
emai
n ho
me,
mob
ile p
hone
s en
able
d ra
pid
acce
ss t
o cl
inic
al a
nd s
ocia
l sup
port
net
wor
ks, P
C s
ervi
ce
(nur
ses
and
volu
ntee
rs)
impo
rtan
t
Serv
ice
eval
uatio
n, s
elec
tion
bias
, sm
all s
ampl
es P
allia
tive
care
wor
ker
pres
ent →
may
in
hibi
t ho
nest
res
pons
es
Med
ium
Han
sen
et a
l.26
USA
To
desc
ribe
the
ex
peri
ence
s of
FC
Gs
with
fo
rmal
and
info
rmal
car
e at
the
end
-of-l
ife fo
r dy
ing
olde
r ad
ults
Rur
al s
tudy
23 F
CG
(be
reav
ed)
Rec
ruitm
ent:
purp
osiv
e sa
mpl
e, id
entif
ied
by
fune
ral d
irec
tor
and
hosp
ice
staf
f. In
vita
tion
mai
led
2–12
mon
ths
afte
r th
e de
ath
Dia
gnos
is: c
ance
r an
d ch
roni
c ill
ness
Rur
al, a
gric
ultu
ral
Paci
fic N
orth
wes
t co
unty
; ave
rage
of 1
1.2
peop
le p
er s
quar
e m
ile
Gen
eric
, qua
litat
ive,
exp
lora
tory
de
sign
; sem
i-str
uctu
red,
ope
n-en
ded
ITV
s; p
urpo
sive
sam
plin
g; t
hem
atic
an
alys
isIn
terv
iew
er: r
esea
rche
r
Rur
al s
ettin
g:Be
nefit
sN
eigh
bour
s, fr
iend
s, v
olun
teer
s of
fere
d ho
useh
old
help
and
res
pite
car
eFo
rmal
car
e pr
ovid
ers
ofte
n kn
own
pers
onal
lyFl
exib
ility
in o
rgan
isat
iona
l sup
port
, mor
e cr
eativ
eSu
ppor
t fr
om c
hurc
h m
embe
rsCh
alle
nges
Lim
ited
reso
urce
s fo
r co
ntin
uity
of c
are
Geo
grap
hica
l ser
vice
bou
ndar
ies →
limite
d av
aila
bilit
y, g
reat
er t
rave
l dis
tanc
esLa
ck o
f kno
wle
dge
abou
t Eo
LC b
y pa
id
care
give
rsFC
G s
tres
s
Smal
l sam
ple
Hig
h
Hat
cher
et
al.2
7
Aus
tral
ia
To
docu
men
t ca
rer
perc
eptio
ns o
f pat
ient
s’
tran
sitio
ns fr
om c
omm
unity
to
hos
pita
l-bas
ed p
allia
tive
care
Rur
al s
tudy
6 FC
Gs
(act
ive)
Rec
ruitm
ent:
on
adm
issi
on t
o PC
ser
vice
Dia
gnos
is: c
ance
r
Rur
al c
omm
unity
; not
de
scri
bed
Qua
litat
ive,
sem
i-str
uctu
red
ITV
s,
first
ITV
at
time
patie
nt t
rans
ferr
ed
to h
ospi
tal,
seco
nd IT
V 3
mon
ths
late
r, c
odin
g, t
hem
atic
ana
lysi
sIn
terv
iew
er: n
on-t
reat
ing
PC s
ervi
ce
staf
f
Tra
nsfe
r to
hos
pita
l occ
urre
d at
a c
risi
s po
int
and
not
diss
atis
fact
ion
with
hom
e-ca
re s
ervi
ce,
expe
rien
ces
mos
tly p
ositi
veD
ecis
ion
to t
rans
fer
initi
ated
and
led
by F
CG
–
inab
ility
to
prov
ide
optim
al c
are;
phy
sica
l di
fficu
lties
, pat
ient
pai
n an
d an
xiet
ySm
ooth
tra
nsiti
on t
o co
mm
unity
hos
pita
l due
to
goo
d co
mm
unic
atio
n an
d co
ntin
uity
of c
are
by P
C n
urse
Hos
pita
l car
e –
mos
tly s
atis
fied;
per
sona
l, tr
eate
d as
indi
vidu
als,
com
pass
iona
te, s
afe
and
conv
enie
nt t
o vi
sit
Neg
ativ
e as
pect
s re
late
d to
pai
n, in
terv
entio
ns,
com
mun
icat
ion
and
sens
e th
at h
ospi
tal i
nfer
red
impe
ndin
g de
ath
Smal
l sam
ple,
sin
gle
PC
serv
ice,
sel
f-sel
ectio
n, c
ance
r,
dist
ance
to
hosp
ital n
ot
repo
rted
Med
ium
Tab
le 1
. (C
ontin
ued)
Rainsford et al. 901
Purp
ose
Info
rman
ts/r
ecru
itmen
tR
ural
def
initi
onM
etho
dolo
gyFi
ndin
gsLi
mita
tions
/bia
sQ
ualit
y
Her
ce e
t al
.28
Mal
awi
To
eval
uate
PC
pro
gram
me
outc
omes
, to
unde
rsta
nd
PC n
eeds
, kno
wle
dge
and
pref
eren
ces
with
the
aim
of
serv
ice
impr
ovem
ent
Rur
al s
tudy
36 p
atie
nts
(incl
uded
9
dyad
s), 1
1 FC
Gs
(act
ive)
Rec
ruitm
ent:
on
adm
issi
on t
o PC
pr
ogra
mm
e tr
eatm
ent
clin
ics
Dia
gnos
is: K
apos
i sa
rcom
a, H
IV, c
ance
r
Nen
o di
stri
ct,
(pop
ulat
ion
130,
612)
is
olat
ed r
ural
are
a
Rap
id e
valu
atio
n m
etho
d (R
EM);
char
t re
view
Stru
ctur
ed a
nd o
pen-
ende
d IT
Vs,
th
emat
ic a
naly
sis
9 jo
int
ITV
sIn
terv
iew
er: S
tudy
inve
stig
ator
s,
inte
rpre
ter
and
HW
pre
sent
Gre
ates
t ne
eds
wer
e fin
anci
al (
inco
me
and
food
), hi
gh p
reva
lenc
e of
pai
n (6
4%
mod
erat
e to
sev
ere)
, 75%
of p
atie
nts
rece
ived
ps
ycho
soci
al s
uppo
rt fr
om fa
mily
Patie
nts’
pre
ferr
ed p
lace
of c
are
hom
e 46
%,
com
mun
ity fa
cilit
y 30
%, h
ospi
tal 2
4%, F
CG
: 46
%, 4
6%, 3
6%FC
G: f
elt
oblig
ated
, som
e vo
lunt
eere
d. T
ime
com
mitm
ent
– A
DLs
, med
icat
ion,
em
otio
nal
supp
ort,
sym
ptom
car
e, t
rans
port
Smal
l stu
dy, s
peci
fic c
ultu
ral
issu
es (
whi
le t
his
stud
y is
ge
nera
lisab
le t
o su
b-Sa
hara
n A
fric
a, it
is n
ot n
eces
sari
ly
gene
ralis
able
in t
he c
onte
xt
of t
his
revi
ew)
Med
ium
John
ston
et
al.2
9
Scot
land
To
unde
rsta
nd p
atie
nt a
nd
care
r ex
peri
ence
s of
EoL
C,
to e
xplo
re s
elf-c
are
copi
ng
stra
tegi
esR
ural
and
rem
ote
stud
y
20 p
atie
nts,
the
ir m
ain
care
r (a
nd t
heir
HC
P)R
ecru
itmen
t: pu
rpos
ive
sam
ple,
PC
com
mun
ity
nurs
eD
iagn
osis
: adv
ance
d ca
ncer
Rur
al a
nd r
emot
e ar
eas
in H
ighl
ands
and
Wes
t of
Sco
tland
Qua
litat
ive;
53
in-d
epth
, un
stru
ctur
ed s
eria
l IT
Vs
(2 w
eeks
ap
art)
, tri
angu
latio
n, t
hem
atic
an
alys
is4
join
t IT
Vs
Inte
rvie
wer
: not
spe
cifie
d
Patie
nt s
elf-c
are
copi
ng s
trat
egie
s in
clud
edM
aint
aini
ng n
orm
ality
(go
al s
ettin
g, t
akin
g a
brea
k, in
depe
nden
ce, r
emai
n ho
me)
Prep
arin
g fo
r de
ath
(wha
t to
exp
ect,
fune
ral
plan
ning
, say
ing
good
bye)
Supp
ort
from
fam
ily a
nd fr
iend
s (p
hysi
cal a
nd
emot
iona
l), s
uppo
rt fo
r fa
mily
oft
en la
ckin
gPh
ysic
al (
sym
ptom
s, fi
nanc
ial,
aids
to
hous
e)Em
otio
nal (
acce
ptan
ce, b
eing
pos
itive
, con
trol
, ch
oice
, hop
e, r
elig
ion,
pee
rs)
Supp
ort
from
HC
P (h
ome
help
, out
of h
ours
ca
re)
Smal
l sam
ple,
rec
ruitm
ent
bias
, can
cer
only
Med
ium
Kel
ly e
t al
.30
Can
ada
To
unde
rsta
nd c
ross
-cu
ltura
l hos
pita
l-bas
ed
end-
of-li
fe c
are
from
the
pe
rspe
ctiv
e of
ber
eave
dFi
rst
Nat
ions
fam
ily
mem
bers
Rur
al s
tudy
10 r
ecen
tly b
erea
ved
abor
igin
al fa
mily
mem
bers
Rec
ruitm
ent:
conv
enie
nce
sam
ple,
live
d ne
ar h
ealth
ce
ntre
whe
re fa
mily
m
embe
r di
edD
iagn
osis
: not
spe
cifie
d
A r
ural
tow
n in
N
orth
ern
Ont
ario
with
a
catc
hmen
t of
23,
000
Ojib
way
and
Cre
e ab
orig
inal
pat
ient
s
Phen
omen
olog
ical
sem
i-str
uctu
red
ITV
s, c
ryst
allis
atio
n an
d im
mer
sion
te
chni
ques
, tri
angu
latio
n an
d m
embe
r-ch
ecki
ng m
etho
dsIn
terv
iew
er: n
urse
s + re
sear
ch
assi
stan
t, no
t in
volv
ed in
car
e
Cro
ss-c
ultu
ral c
are
at t
he t
ime
of d
eath
is
chal
leng
ing
Serv
ice
deliv
ery
and
com
mun
icat
ion
stra
tegi
es
mus
t m
eet
cultu
ral a
nd fa
mily
nee
ds R
espe
ct
(res
pect
ful d
irec
tnes
s), c
omm
unic
atio
n,
appr
opri
ate
envi
ronm
ents
(e.
g. s
pace
to
acco
mm
odat
e a
larg
er n
umbe
r of
vis
itors
) an
d ca
regi
ving
wer
e im
port
ant
to p
artic
ipan
ts fo
r cu
ltura
lly a
ppro
pria
te p
allia
tive
care
Firs
t N
atio
ns fa
mily
mem
bers
des
crib
ed
palli
ativ
e ca
re a
s a
com
mun
ity a
nd e
xten
ded
fam
ily e
xper
ienc
eG
ener
ally
, rel
atio
nshi
ps w
ith n
urse
s an
d th
e ca
re t
he n
urse
s pr
ovid
ed w
ere
posi
tive
expe
rien
ces
Smal
l sam
ple,
sin
gle
serv
ice,
sp
ecifi
c cu
lture
, onl
y En
glis
h-sp
eaki
ng fa
mili
es r
ecru
ited
Med
ium
Tab
le 1
. (C
ontin
ued)
(Con
tinue
d)
902 Palliative Medicine 31(10)
Purp
ose
Info
rman
ts/r
ecru
itmen
tR
ural
def
initi
onM
etho
dolo
gyFi
ndin
gsLi
mita
tions
/bia
sQ
ualit
y
Lock
ie e
t al
.31
Pesu
t et
al.3
2T
o in
vest
igat
e th
e ex
peri
ence
s of
rur
al c
ance
r pa
tient
s an
d th
eir
FCG
s,
who
com
mut
e to
an
urba
n ca
ncer
cen
tre
for
palli
ativ
e ca
reR
ural
stu
dy
15 p
atie
nts
and
thei
r FC
Gs
Rec
ruitm
ent:
atte
ndin
g re
gion
al c
ance
r ce
ntre
, id
entif
ied
by r
esea
rche
r,
appr
oach
ed b
y vo
lunt
eer
Dia
gnos
is: a
dvan
ced
canc
er
Sout
h ce
ntra
l reg
ion
of B
ritis
h C
olum
bia,
ru
ral a
nd r
emot
e co
mm
uniti
es, o
utsi
de
a m
ajor
urb
an c
entr
e,
popu
latio
n <
10,0
00,
mou
ntai
ns w
ith w
inte
r sn
ow a
nd ic
ePo
pula
tion
dens
ity 3
.4
peop
le/k
m2 ,
aver
age
leng
th o
f com
mut
e 17
7 km
(on
e w
ay)
Mix
ed m
etho
ds: s
emi-s
truc
ture
d IT
Vs
and
wri
tten
que
stio
nnai
re,
codi
ng, i
tera
tive
proc
ess,
the
mes
and
su
bthe
mes
Inte
rvie
wer
: res
earc
h as
sist
ant
Bene
fits
of r
ural
livi
ng: e
nvir
onm
ent,
priv
acy,
co
mm
unity
sup
port
, fri
ends
, fun
drai
sers
. In
tang
ible
s of
goo
d he
alth
-hap
pine
ss,
cont
entm
ent,
peac
eCh
alle
nges
of c
omm
utin
g lo
ng d
ista
nces
(w
eath
er, r
oad
cond
ition
s, b
reak
dow
ns, a
cces
s to
mob
ile p
hone
cov
erag
e), c
osts
of f
uel a
nd
acco
mm
odat
ion,
tra
vel w
as a
dded
fina
ncia
l bu
rden
(co
sts)
and
tim
e lo
st (
wor
k, fa
mily
and
co
mm
unity
res
pons
ibili
ties)
Posi
tives
of a
tten
ding
can
cer
cent
re –
co
mpa
ssio
nate
car
e, e
ncou
rage
men
tC
omm
utin
g: ‘m
akin
g it
wor
k’Pl
anni
ng a
nd p
repa
ratio
n, w
ell o
rgan
ised
, su
ffici
ent
pain
med
icat
ion,
rou
tine,
tim
ing
of
appo
intm
ents
Dea
ling
with
unf
amili
ar t
erri
tory
The
tol
l of c
omm
utin
g –
finan
cial
, tim
e, F
CG
s’
own
heal
thM
akin
g th
e be
st o
f it
– ut
ilisi
ng t
ime
wel
l (e
rran
ds, s
hopp
ing,
vis
it fr
iend
s)
Smal
l sam
ple
size
, can
cer
patie
nts
Hig
h
Mix
er e
t al
.33
USA
To
desc
ribe
gen
eric
(fo
lk)
fact
ors
that
hea
lthca
re
wor
kers
can
app
ly
to p
rovi
de c
ultu
rally
co
ngru
ent
end
of li
fe c
are
Rur
al s
tudy
6 ke
y in
form
ants
(pa
tient
s w
ith li
fe-li
miti
ng il
lnes
s an
d/or
the
ir F
CG
(a
ctiv
e +
ber
eave
d)) =
11
part
icip
ants
in t
otal
(9 g
ener
al in
form
ants
(H
CP/
past
or))
Rec
ruitm
ent:
hosp
ice
wor
kers
and
pas
tors
Dia
gnos
is: L
ife-li
miti
ng
illne
ss
Rur
al s
ettin
g –
East
T
enne
ssee
reg
ion
of r
ural
App
alac
hia
– a
regi
on m
edic
ally
un
ders
ervi
ced,
with
19
% li
ving
bel
ow t
he
pove
rty
line
Qua
litat
ive
ethn
onur
sing
m
etho
dolo
gy, s
emi-s
truc
ture
d op
en-
ende
d IT
Vs,
the
mat
ic a
naly
sis
Inte
rvie
wer
: pri
ncip
al in
vest
igat
or
and/
or t
rain
ed s
tude
nts
Dig
nity
thr
ough
res
pect
of f
olk
cust
om, b
elie
fs
and
faith
and
inte
grat
ion
of fo
lk c
are
(e.g
. fol
k re
med
ies)
into
EoL
CSp
iritu
al c
are
high
ly v
alue
d, fa
ith fu
ndam
enta
l to
tra
nsiti
on t
hrou
gh E
OL
Div
ersi
ty in
sup
port
pr
ovid
ed b
y ch
urch
mem
bers
Fam
ily c
are
at E
OL
esse
ntia
l for
cul
tura
lly
cong
ruen
t ca
re –
at
hom
e, q
ualit
y fa
mily
tim
e,
how
ever
, bur
den
of c
are →
fam
ily c
ohes
ion
or
conf
lict
Expe
ct H
CP
to d
emon
stra
te t
rust
ing
and
cari
ng
beha
viou
rN
o pa
edia
tric
pal
liativ
e ca
re s
ervi
cePa
lliat
ive
care
oft
en m
isun
ders
tood
Sym
ptom
m
anag
emen
t ac
cept
ed, h
owev
er, s
ervi
ce
ofte
n de
clin
ed d
ue t
o be
lief P
C is
con
trar
y to
co
ntin
uing
cur
ativ
e ca
re o
r as
soci
ated
with
st
igm
a of
dea
thSo
me
patie
nts
pref
erre
d to
sta
y in
hos
pita
l tha
n re
turn
hom
e
Smal
l sam
ple
size
, spe
cific
cu
ltura
l gro
up w
ithin
sp
ecifi
ed r
egio
n, s
elf-
repo
rted
as
Chr
istia
n,
sele
ctio
n bi
as
Med
ium
Tab
le 1
. (C
ontin
ued)
Rainsford et al. 903
Purp
ose
Info
rman
ts/r
ecru
itmen
tR
ural
def
initi
onM
etho
dolo
gyFi
ndin
gsLi
mita
tions
/bia
sQ
ualit
y
Ost
erta
g an
d Fo
rem
an34
USA
To
desc
ribe
bas
elin
e co
mm
unity
dem
ogra
phic
s an
d re
sour
ces,
to
star
t a
com
mun
ity in
quir
y in
to c
once
rns
at E
oL b
y ex
plor
ing
the
opin
ions
and
ex
peri
ence
s of
com
mun
ity
mem
bers
and
HC
PsR
ural
stu
dy
19 fa
mily
mem
bers
of
patie
nts
who
had
die
d w
ith (
13)
or w
ithou
t (6
) ho
spic
e se
rvic
es.
(53
HC
Ps, 9
hos
pice
vo
lunt
eers
)D
isea
se: >
50%
can
cer
or
hear
t di
seas
e
Han
cock
cou
nty,
rur
al
com
mun
ityFo
cus
grou
ps, t
rans
crib
ed, t
hem
atic
an
alys
isFa
mily
mem
ber’
s pa
in, p
hysi
cal s
ympt
oms
and
anxi
etie
s w
ell-m
anag
ed r
egar
dles
s of
hos
pice
or
non-
hosp
ice
stat
usPo
or p
hysi
cian
com
mun
icat
ion
com
mon
Man
y no
n-ho
spic
e FC
G e
xpre
ssed
the
wis
h th
at h
ospi
ce-t
ype
serv
ices
had
bee
n of
fere
d to
the
m, f
or e
xam
ple,
ass
ista
nce
to fa
cilit
ate
a ho
me
deat
hT
hose
not
enr
olle
d w
ere
mai
nly
chro
nic
non-
canc
er p
atie
nts;
fam
ilies
wer
e un
prep
ared
for
term
inal
pha
se a
nd d
eath
App
reci
ated
pri
mar
y ca
re p
hysi
cian
s w
ho
mai
ntai
ned
a pr
esen
ce u
ntil
the
end,
som
e w
ere
‘invi
sibl
e’ o
nce
cure
no
long
er p
ossi
ble
Tho
se w
ho h
ad u
sed
hosp
ice
wer
e ap
prec
iativ
e an
d w
ould
rec
omm
end
it to
oth
ers
Smal
l sam
ple,
sin
gle
coun
tyM
ediu
m
Payn
e et
al.1
1
UK
To
desc
ribe
the
ex
peri
ence
s of
pat
ient
s an
d ca
rers
of E
oLC
in
com
mun
ity h
ospi
tals
Mix
ed g
eogr
aphi
cal s
tudy
18 e
lder
ly p
atie
nts
(⩾65
year
s) a
nd t
heir
FC
Gs
(n =
11)
Rec
ruitm
ent:
hosp
ital
nurs
esD
iagn
osis
: can
cer
or
adva
nced
con
ditio
n
4/6
site
s w
ere
rura
l3
smal
l GP-
led
and
1 m
ediu
m s
hare
d –
care
ho
spita
ls in
sm
all t
owns
or
rur
al a
reas
Sem
i-str
uctu
red
ITV
s (8
join
t IT
Vs)
, gr
ound
ed t
heor
y, o
pen
codi
ng,
them
atic
ana
lysi
s8
join
t IT
Vs
(5 a
t ho
me
afte
r re
spite
, 3
in h
ospi
tal)
Inte
rvie
wer
: res
earc
her,
rel
atio
nshi
p to
par
ticip
ants
not
spe
cifie
d
Rur
al c
omm
unity
hos
pita
lsFl
exib
ility
Loca
l (ea
sy v
isiti
ng)
Pers
onal
car
eC
omm
unity
hos
pita
ls a
re a
ccep
tabl
e fo
r Eo
LC
Smal
l sam
ple,
rec
ruitm
ent
bias
, 8 jo
int
ITV
s, 2
/3 w
ere
retir
ees
not
‘rur
al lo
cals
’
Hig
h
Pesu
t et
al.3
5,36
Can
ada
To
gain
und
erst
andi
ng
of t
he v
alue
s in
form
ing
and
resp
onsi
bilit
ies
that
su
ppor
t go
od p
allia
tive
care
fr
om r
ural
par
ticip
ants
’ pe
rspe
ctiv
esR
ural
stu
dy
95 r
ural
res
iden
ts
incl
udin
g 25
FC
Gs
iden
tifie
d by
‘cha
mpi
ons’
Rec
ruitm
ent:
purp
osiv
e/s
now
balli
ng, i
dent
ified
by
com
mun
ity P
C c
ham
pion
sD
iagn
osis
: not
spe
cifie
d
4 ru
ral c
omm
uniti
es
in W
este
rn C
anad
a po
pula
tion <
10,0
00,
loca
ted
at le
ast
3 h
by c
ar fr
om s
peci
alis
t pa
lliat
ive
care
tr
eatm
ent
cent
re
Ethn
ogra
phy
51 d
ays
of fi
eld
wor
k,
over
2 ye
ars,
95
sem
i-str
uctu
red
ITV
s an
d 74
h o
f dir
ect
part
icip
ant
obse
rvat
ion.
The
mat
ic a
naly
sis
Inte
rvie
wer
: uni
vers
ity s
tude
nts
Cor
e va
lues
of
Fam
ilies
exp
erie
nced
mea
ning
in c
arin
g w
ith s
atis
fact
ion
outw
eigh
ing
the
burd
en;
resp
onsi
bilit
ies
incl
ude
prov
idin
g, m
anag
ing
and
co-o
rdin
atin
g ca
re a
nd a
dvoc
acy
Pain
and
sym
ptom
man
agem
ent
a pr
iori
ty w
ith
expe
ctat
ion
prof
essi
onal
sup
port
ava
ilabl
e. T
he
degr
ee o
f sym
ptom
con
trol
sha
ped
perc
eptio
n of
qua
lity
of d
ying
Kno
win
g an
d be
ing
know
n –
adva
ntag
es a
nd
disa
dvan
tage
s as
illn
ess
traj
ecto
ry a
nd g
rief
w
ere
publ
ic e
vent
s w
ith lo
ss o
f ano
nym
ityBe
ing
pres
ent
and
avai
labl
e –
FCG
nee
ded
on-
site
sup
port
not
just
a p
hone
cal
lC
omm
unity
and
mut
ualit
y es
peci
ally
tan
gibl
e su
ppor
t su
ch a
s fu
ndra
isin
g, h
ome
reno
vatio
ns,
mea
ls. O
ften
tho
se p
artic
ipan
ts w
ho w
ere
‘giv
ers’
to
the
com
mun
ity r
ecei
ved
the
high
est
amou
nt o
f sup
port
Fam
ilies
, HC
P an
d ad
min
istr
ator
s ne
ed t
o w
ork
toge
ther
Rem
ote
heal
thca
re d
ecis
ions
can
sev
erel
y un
derm
ine
loca
l cap
acity
cau
sing
frag
men
tatio
n in
alr
eady
und
erse
rvic
ed c
omm
uniti
es
Sing
le r
egio
n, s
now
balli
ng
recr
uitm
ent
bias
, no
indi
catio
n if
FCG
s ar
e ac
tive
or le
ngth
of b
erea
vem
ent
Hig
h
(Con
tinue
d)
Tab
le 1
. (C
ontin
ued)
904 Palliative Medicine 31(10)
Purp
ose
Info
rman
ts/r
ecru
itmen
tR
ural
def
initi
onM
etho
dolo
gyFi
ndin
gsLi
mita
tions
/bia
sQ
ualit
y
Rev
ier
et a
l.37
USA
To
expl
ore
the
mea
ning
of
hop
e fo
r th
e FC
G in
th
e co
ntex
t of
end
-of-l
ife
care
, ide
ntify
whi
ch n
ursi
ng
actio
ns in
fluen
ce h
ope
Rur
al s
tudy
FCG
(n
= 6
); cu
rren
tly
cari
ng o
r ha
d ca
red
with
in
prev
ious
6 w
eeks
Rec
ruitm
ent:
iden
tifie
d by
hos
pice
pro
gram
me
nurs
es r
etur
n co
nsen
t by
m
ail t
o re
sear
cher
Dia
gnos
is: n
ot s
peci
fied
Rur
al, M
idw
est
USA
Phen
omen
olog
ical
; tw
o tim
e-po
int
dial
ogic
al e
ngag
emen
t (h
ospi
ce e
nrol
men
t an
d 4–
6 w
eeks
be
reav
emen
t); d
ata
anal
ysis
: ex
trac
tion-
synt
hesi
s, h
euri
stic
in
terp
reta
tion
Inte
rvie
wer
: res
earc
her
Four
inte
rcon
nect
ing
them
es a
s fo
llow
s:En
gagi
ng, s
tren
gthe
ning
and
mai
ntai
ning
co
nnec
tions
Easi
ng o
f sel
fFi
ndin
g m
eani
ng in
the
situ
atio
nA
ccep
tanc
eH
ope
for
FCG
ass
ists
nav
igat
ion
thro
ugh
the
care
givi
ng e
xper
ienc
e, fi
ndin
g po
sitiv
e m
eani
ng
with
in t
he s
ituat
ion
and
mov
ing
tow
ards
ac
cept
ance
of t
he s
ituat
ion
and
heal
ing
Hop
e su
ppor
ted
thro
ugh
rela
tions
hip
with
nu
rses
and
info
rmat
ion
Smal
l sam
ple,
gat
ekee
ping
, re
crui
ted
by n
urse
sM
ediu
m
Tam
anna
i et
al.3
8
Cam
eroo
n
To
unde
rsta
nd t
he n
eeds
of
patie
nts
and
thei
r fa
mili
es
visi
ted
by a
chi
ldre
n’s
PC n
urse
in C
amer
oon
and
to id
entif
y as
pect
s of
th
e se
rvic
e th
at c
an b
e im
prov
edR
ural
stu
dy
3 pa
tient
s (1
4, 9
and
10
year
s); 7
FC
Gs
(3
mot
hers
, 2 fa
ther
s,
1 gr
andm
othe
r, 1
gr
andf
athe
r)R
ecru
itmen
t: no
t re
port
edD
iagn
osis
: pae
diat
ric
Burk
itt ly
mph
oma
FGC
occ
upat
ion
– ‘fa
rmer
’Q
ualit
ativ
e, s
emi-s
truc
ture
d IT
Vs
with
ope
n-en
ded
Q, p
rede
sign
ed
topi
c gu
ide,
the
mat
ical
ly a
naly
sed
Inte
rvie
wer
: not
spe
cifie
d
Nee
ds id
entif
ied:
fina
ncia
l aid
, gen
eral
dis
ease
im
prov
emen
t an
d pr
ayer
sC
arer
s di
d no
t kn
ow w
hat
to e
xpec
t at
hom
e an
d ho
w t
o m
anag
e pr
even
tabl
e pr
oble
ms
Inco
ngru
ence
in F
CG
s’ a
nd p
atie
nts’
vie
ws
on
clin
ical
sta
tus
and
need
sO
pen
com
mun
icat
ion
betw
een
FCG
s an
d pa
tient
s w
as c
halle
ngin
gFa
ilure
to
cure
→ ac
cept
ance
out
com
e w
as in
G
od’s
han
dsT
he d
isea
se g
reat
ly in
fluen
ced
QoL
of p
atie
nts
and
FCG
s
Smal
l sam
ple,
sin
gle
serv
ice
recr
uitm
ent,
prog
ram
me
eval
uatio
n, R
R n
ot p
rovi
ded,
bi
as a
s al
l rec
eivi
ng fr
ee P
C
(inte
rvie
wee
cou
rtes
y bi
as),
limite
d to
lym
phom
a
Hig
h
Will
iam
s et
al.3
9
Can
ada
To
shar
e, in
the
form
of a
st
ory,
the
exp
erie
nces
of
rura
l fem
ale
FCG
s fo
cusi
ng
on w
hat
fost
ers
thei
r ho
pe(L
ivin
g w
ith H
ope
Prog
ram
)R
ural
stu
dy
23 fe
mal
e FC
Gs
(act
ive)
Rec
ruitm
ent:
conv
enie
nce
sam
ple,
initi
al c
onta
ct b
y PC
ser
vice
via
mai
lD
iagn
osis
: adv
ance
d ca
ncer
Livi
ng o
utsi
de m
ajor
po
pula
tion
area
s in
Sa
skat
chew
an a
nd
Alb
erta
, rur
al a
reas
de
sign
ated
by
prov
inci
al
post
al c
odes
Dai
ly 5
-min
jour
nal e
ntry
on
chal
leng
es a
nd h
opes
face
d an
d w
hat
fost
ers
hope
; 2-w
eek
peri
od.
Nar
rativ
e en
quir
y ap
proa
ch –
jour
nal
entr
ies
tran
scri
bed
into
nar
rativ
e,
them
atic
ana
lysi
s, r
epor
ted
as a
na
rrat
ive
stor
y ‘H
ope
agai
nst
hope
’
342
jour
nal e
ntri
es,
4 th
emes
:H
ope
‘I ho
pe t
omor
row
is b
ette
r’‘H
opin
g ag
ains
t ho
pe’ →
tens
ion
betw
een
hopi
ng fo
r an
d re
cogn
isin
g th
ere
is n
o cu
reH
ope
is a
cho
ice
and
a m
inds
etH
ope
influ
ence
d by
tem
pora
l cir
cum
stan
ces
(tra
velli
ng, f
inan
ces
moo
d an
d he
alth
of
patie
nt),
soci
al s
uppo
rt; f
aith
and
spi
ritu
ality
Prac
tical
and
em
otio
nal c
halle
nges
(fe
ar, w
orry
, sa
dnes
s, g
uilt,
hel
ples
snes
s, a
nger
, lon
elin
ess,
em
path
y, lo
ve a
nd g
ratit
ude)
Mul
tiple
sel
f-car
e st
rate
gies
The
em
otio
nal j
ourn
ey
Prog
ram
me
eval
uatio
n,
smal
l sam
ple
size
, no
rand
omis
atio
n or
con
trol
gr
oup;
can
cer
only
Des
crip
tive
stor
ies
from
onl
y fe
mal
e FC
Gs
Med
ium
FCG
: fam
ily c
areg
iver
; PC
(S):
palli
ativ
e ca
re (
serv
ice)
; RR
: res
pons
e ra
te; C
BS-E
oL: C
areg
iver
’s B
urde
n Sc
ale
in E
nd-o
f-Life
Car
e; M
SPSS
: Mul
tidim
ensi
onal
Sca
le o
f Per
ceiv
ed S
ocia
l Sup
port
; EC
OG
: Eas
tern
Col
labo
rativ
e O
ncol
ogy
Gro
up p
erfo
rman
ce s
cale
; NO
K: n
ext
of k
in; Q
oDD
: qua
lity
of d
eath
and
dyi
ng; G
AD
-7: g
ener
al a
nxie
ty d
isor
der
7-ite
m s
cale
; PH
Q-8
: pat
ient
hea
lth q
uest
ionn
aire
8-it
em d
epre
ssio
n sc
ale;
EoL
(C):
end
of li
fe (
care
); H
CP:
hea
lthca
re
prof
essi
onal
; ESR
D: e
nd-s
tage
ren
al d
isea
se; C
HF
chro
nic
hear
t fa
ilure
; CO
PD: c
hron
ic o
bstr
uctiv
e pu
lmon
ary
dise
ase;
ITV
: int
ervi
ew; H
NC
: hom
e nu
rsin
g ca
re; A
DLs
: act
iviti
es o
f dai
ly li
ving
; HW
: hea
lth w
orke
r; Q
oL: q
ualit
y of
life
; SD
: sta
ndar
d de
viat
ion;
CI:
conf
iden
ce in
terv
al.
Tab
le 1
. (C
ontin
ued)
Rainsford et al. 905
Figure 1. PRISMA flowchart.Source: Modified flow chart as described by Moher et al.40
FCG: family caregiver; EoLC: end-of-life care.
and one study included family caregivers of patients receiving or not receiving hospice care.34 Length of end-of-life care by family caregivers ranged from 2 weeks to 120 months;26,28 11 studies did not report length of care.
Patient experiences and perspectives
In total, 11 studies described the end-of-life care experi-ences and perspectives of rural patients and emphasised
the importance of not giving up,16,19–21 finding meaning in life,21 ‘steadfastly living life’,16 maintaining dig-nity,16,19,25,29 independence,16,19,21–24,29 and normality.16,19,29 This required patients to redefine normal,16,19,20,23,29 come to terms with change,21,23,29 (with resignation16 or struggle19) and make the most of everyday.16 Patients lived with exhaustion and stress,19,20,22,23 ‘in conflicting states of hope and despair (p. 783)’,19 and balanced inde-pendence with an awareness of their deterioration and
906 Palliative Medicine 31(10)
Table 2. Characteristics of patients.
Patient (n=) 167Studies including patients (n=) 12Gender Female 81 Male 53 Not specified 33Age (n studies; range) Range 9–93 years Paediatrics 1 (9–14 years) Adults 5 (30–93 years) All >65 years 4 Age not reported 2Diagnosis (n studies) Cancer only 8 Non-cancer only 0 Cancer + non-cancer 3 Not specified 1Receiving PC (n studies) Yes 10 Interviewed at time of enrolment into
PC service1
Not specified; attending oncology outpatients
1
PC: palliative/hospice care.
Table 3. Characteristics of FCGs.
Studies including FCG (n=) 19FCG (n=) 1086 Active 280 (26%) Bereaved 770 (71%) Not specified 36 (3%)Gender Female 762 (70%) Male 227 (21%) Not specified 97 (9%)Age Range 27–90 years Mean age >60 years 4 studies Age not reported 9 studiesRelationship to patient Spouse/partner 422 (39%) Parent 12 (1%) Adult child 402 (37%) Sibling 5 (0.4%) Other family 237 (22%) Friend 8 (0.6%)Recipient enrolled in PC Yes 17 studies Not specified 1 study Interview at time of enrolment in PC service 1 study
FCG: family caregiver; PC: palliative/hospice care.
the future;19,20,29 and endured life bravely while having no energy left to enjoy life.19,20
Dignity was maintained by refusing to be defined by illness16,19,22,23,29 and finding hope19,21,23,25,29 as ‘hope is the key to enduring distress (p. 785)’.19 Awareness of increas-ing dependence on family and friends was associated with fear of becoming a burden16,19,22,23,29 and/or losing inde-pendence, so often they did not ask for help.19,22 One Norwegian patient found hope in pursing life-prolonging chemotherapy despite losing dignity through side effects,19 while others refused chemotherapy in order to maintain their quality of life.29
Preparing for death, ‘without loss of hope or the desire to keep living (p. 1622)’29 was important for many patients. Patients engaged in funeral planning,29 complet-ing wills,16,29 preparing self and family for impending separation,16,29 and setting affairs in order.16,18
In sub-Saharan Africa, pain dominated the lives of patients25,28 with 31 (86% of patients) in Malawi reporting pain to be moderate to severe.28 Pain issues were reported in four studies from developed countries.20,21,29,32 with participating patients describing pain management as important,21 essential for self-care,29 and the importance of anticipating medication requirements when commuting long distances.31,32
Support of family, friends, community,22,28 and HCPs23 was seen as essential and valued; however, despite this connection, the message from two studies was that patients felt isolated as their disease progressed, ‘I’m part of the community but I feel alone (p. 2)’22 and having to ‘walk the palliative path alone (p. 12)’.20
Family caregiver experiences and perspectives
Rural family caregivers spoke of taking on the responsibil-ity of providing ‘direct care, managing and coordinating care, and advocacy (p. 127)’36 for their family member; however, the care provided was ‘not only about the dying person, but also about the living relatives (p. 392)’.18 Some family caregivers reported lacking knowledge29 and with ‘few [being] physically, emotionally, or educationally pre-pared for the tasks and responsibilities of caregiving (p. 5)’37 especially as the illness progressed.22 Family car-egivers experienced a broad spectrum of negative emo-tions.39 ‘Participants were very concerned with ensuring the dignity and comfort of their loved ones during their final days and experienced distress, guilt, and anger if they were unable to fulfil these self-designated responsibilities (p. 6)’.37 If family caregivers could find meaning in the situation by focusing hope on the day-to-day moments,37 redefining normal,23,39 connecting and separating,37 they were better positioned to cope with the responsibilities, burden of care and transitions in roles.23 Self-care and maintaining their own quality of life were important; how-ever, Williams reported that family caregivers often lacked
increasing dependence on others;22,29 talked about impending death while holding onto hopes and dreams for
Rainsford et al. 907
the ‘time and energy to do everything they needed and to care for themselves (p. 6)’.39
To fulfil their caregiver responsibilities, family car-egivers required support14,15,22,28,35,36 from family, friends, neighbours and HCPs and they ‘identified the need of having someone to talk to and being appreciated by the care recipient (p. 15)’.14 Some family caregivers accepted the role out of ‘familial obligation’;28 however, many found meaning in caring for their loved ones36,37 with the burden of care ‘outweighed by … the satisfaction they derived from having made a meaningful contribution (p. 130)’.36
Facilitators and barriers to receiving end-of-life care in rural/remote settings
Communication; accurate and timely information. The greatest support need of participating rural patients and family caregivers in developed countries was informa-tional; however, the need for accurate information was stated or implied in all included studies from both devel-oped and developing countries. Effective communication between HCPs and patients/family caregivers and within families38 reduced pain and distress,25 empowered carers to fulfil their responsibilities,37 facilitated smooth transi-tions of care,27 and allowed patients and families to pre-pare for death.18,29 Most, but not all19 participants were satisfied with the standard of communication by rural HCPs. Participants in nine studies11,14,15,17,19,20,22,23,27,29,34,38 reported one or more communication difficulties such as receiving conflicting or untimely information,34,36 uncer-tainty as to ‘which physician was in charge’,17 and not receiving information from the person they considered to be the expert.23,36 While some patients accepted medical advice without questioning,16 others considered false hope30 or poor communication a lack of respect for the patient and their family22,23,30 and that ‘not knowing was worse than knowing (p. 1623)’.29 One study revealed that primary care physicians were highly praised for honesty and presence at the time of death.34
Formal services. Brazil et al.14 reported 82.6% of participat-ing rural family caregivers indicated that formal palliative care services were readily available, with 68.6% having access to services after hours. While another study35 reported a HCP was always available, these results were not universal with four qualitative studies reporting inad-equate accessibility to care and continuity of care19,20,22,26,29 especially after hours.22,23,26,29 Access to HCPs with pallia-tive care training,23,26 paid qualified in-home carers,26 after-hours pharmacies or morphine,25,26 respite care14,15 and paediatric hospice33 was limited or unavailable.
Features of care that facilitated quality end-of-life care included personalised care;23,26 knowing and being known by the HCPs;35 and a willingness of HCPs to go beyond
their professional care.35 However, loss of privacy and anonymity35 and an expectation that friends will always be available35 were perceived as barriers. Five studies19–23,26,32 reported that the quality of care provided was dependent on the personality of the HCP with difficulties arising if personality conflicts arose as often no alternative provider was available.19
In three studies, when care at home was not possible, the local community hospital was an acceptable alternative26,28,33 and were considered safe,26 small, convenient, personal, wel-coming and the nursing staff described as caring and com-passionate. However, rural hospitals were not viewed positively by Indigenous participants in Australia18 and Canada,30 mainly due to cultural insensitivity.
Informal social support – family, friends and neighbours. A total of 19 studies reported on the informal support pro-vided by family, friends, neighbours and the community with some participants stating that family is the ‘most important’ factor19,21 and essential for ‘culturally congru-ent care’.33 Brazil et al.14,15 reported that the greatest unmet needs identified by rural family caregivers were the tangi-ble or practical needs. Community support was reported to have a positive influence on rural end-of-life care with one participant describing a sense of solidarity as ‘[we] take care of each other … That’s just the way it is! (p. 5)’21 It was acknowledged in two studies that not all patients have happy family relationships33,39 and in another that commu-nity support could not be taken for granted and was highly reciprocal in that those participants who had been involved in giving to their rural community also received the high-est amount of support from that community.35 Despite the strong sense of community, studies reported that as disease progressed and patients lost mobility and independence, there was a sense of isolation as quoted from a study par-ticipant: ‘I am part of my community but I feel alone. Fam-ily and friends come to visit me, but I feel isolated as they are unable to understand what is happening to me and my wife (p. 2)’.22
Emotional support. Strong emotional support was identi-fied as a facilitator of quality rural end-of-life care and was dependent on good communication, information, the presence of HCPs, support of other patients,29 faith and hope. Hope was maintained through connection with family,19 friends and being linked to something outside the illness.37
Spiritual support. Spiritual connection and faith fostered hope,30,33,37,39 with faith seen as an enabler to persevering in life as death drew near. Faith was reported to be funda-mental to rural Appalachians and their transition through end-of-life care.33 In many rural communities, church sup-port was not limited to spiritual issues as congregations also provided physical and financial support.38
908 Palliative Medicine 31(10)
Sub-Saharan Africa. Support needs across the studies were similar; however, contexts varied and the experiences of patients and family caregivers were dependent on where they lived. Three studies25,28,38 were conducted in four developing countries in sub-Saharan Africa where ‘pov-erty shaped how people died (p. 5)’.25 The greatest needs of patients and family caregivers in sub-Saharan Africa were pain relief and access to basic medications,28 practi-cal support, funds to purchase even the basics of life (food, clothing),25,28,38 information regarding the diagnosis and what to expect,25,38 and access to trained HCPs. These three studies were all programme evaluations, and they described how the implementation of palliative care ser-vices improved the quality of life of both patients and fam-ily caregivers by restoring dignity through ‘transforming a life of pain and hopelessness’.25
Influence of rural place and culture
The rural location of participants was important as a cul-tural dimension and ‘participants spoke eloquently of the benefits of their rural lifestyle including physical beauty, privacy and accessibility of recreational activities. The level of support provided by community members was an important factor in why individuals valued rural life (p. 190)’.32
Despite diversity in rural settings each was seen as hav-ing positive and negative influences on rural end-of-life care for both patients and family caregivers with distance identified as the greatest negative influence. One study31,32 focused on the issues surrounding commuting for treat-ment, with another seven studies18–23,25,26,35,36 reporting the experiences relating to travel distances, not only to access outpatient care but also for HCPs in providing home care. Commuting for treatment was mostly seen as stressful and exhausting;19,22,31 inconvenient19 and expensive;25,26,31,32 impacting negatively on the health of family caregivers31 and resulting in fragmented care.22 Some accepted com-muting21 as ‘one of the compromises they have to make for living at home, that is, to live in a place that contributed to their overall health (p. 12)’.20
Geographical distance or ‘living off the beaten track’21 limited accessibility to home-based services as some patients lived outside the boundary for home visits,26 visits were less frequent especially in bad weather18,31,32 and were often not available at short notice or after hours.22,23,26 However, opinions regarding the effect of distance on the quality of care were divided with some seeing it as a ‘major obstacle in providing adequate home-based palliative care (p. 391)’18 and others not viewing rural living as a disad-vantage.21 However, with advanced illness, the partici-pants’ sense of solitude became one of isolation.22 Geographic isolation also explained the greater unmet emotional needs of rural caregivers as they lacked ‘having the support of a group of people who are experiencing the same thing (p. 16)’.14
Three studies reported that mobile phones,25 computers and Internet access21,32 helped reduce the sense of rural isolation by maintaining contact with distant family and improving access to HCPs; however, these technologies are not available everywhere31,32 and for some patients a phone call was not sufficient, as there was a preference for the physical presence of HCPs.26,35
Meaning of home or home country. In one study, 46% of patients and family caregivers reported their preferred place of care to be ‘home’,28 as ‘being at home is like a brick being in the right place: this is my land and these are my people (p. 7)’.21 Many of the rural participants were entrenched in their community and had ‘memories of the landscape, environment, and people they once knew (p. 5)’21 resulting in a strong place attachment (physical, social and autobiographical). It was suggested,
that the rural context may provide an advantageous healthcare environment. Its potential to be a source of comfort, security, and identity concurs with cancer patients’ strong desire for being seen as unique persons … [and a] confidence ‘this place and these people will be there for me’.21 (p. 8)
For indigenous rural residents, ‘home’ or ‘country’ had special cultural significance.18 The biggest barrier to using hospital-based palliative care services for these partici-pants was not being able to die ‘in country’.18 The lack of cultural awareness by HCPs and misperceptions of the concept of palliative care were barriers to accepting pallia-tive care.18,25
Discussion
This systematic literature review describes rural end-of-life care through the experiences and perspectives of rural patients and their family caregivers and illustrates the importance of listening ‘to those experiencing terminal ill-ness and [to] hear what they emphasise as they reflect on their lives (p. 782)’.19 Rural residents clearly hold ‘distinct views’41 on the realities of rural life and rural dying, including benefits and challenges. The voice of rural patients and family caregivers helps ensure services are ‘relevant to [and] embraced by community members (p. 462)’.3 Consistent with previous reports9 rural partici-pants in this literature review were mostly satisfied with the end-of-life care provided to them; however, most were realistic and openly acknowledged their unmet needs which were often ‘related to context … and shaped by reduced access and availability of services’.42 While palli-ative care was available in all the countries included in this systematic review, the development and integration of palliative care into mainstream health service provision within each location were variable. A global mapping study by Lynch et al.43 reported that the provision of pallia-tive care in 2011 remained localised in Cameroon while becoming increasingly integrated in Kenya and Malawi.
Rainsford et al. 909
On a country level, palliative care was at an advanced stage of integration in Uganda and the developed world; however, participants in this systematic review indicated that access remained limited especially in remote areas.
The over-riding themes for most patients and family caregivers in all locations were ‘living life’; holding onto hope, dignity and meaning; receiving personalised care; being known; and for HCPs to demonstrate ‘presence, reassurance and honouring choices’.44 In sub-Saharan Africa, this was possible once pain was managed. The importance of family and dying within one’s community was expressed by most participants; however, this was more significant for indigenous participants who consid-ered cultural sensitivity and respect for their rituals18 and ‘folk culture’33 to be essential for end-of-life care. The end of life issues faced in general by all patients and family caregivers, regardless of where they live,45–47 were raised by the rural participants in this review; however, there are differences, facilitators and barriers, unique to rural set-tings that significantly impact rural end of life care.
Barriers to providing rural end of life care, such as the hardship of distance and isolation, are not only just the concerns of rural patients and family caregivers but also acknowledged in the literature by rural palliative and community nurses.48,49 While there is no expectation that resources in rural areas should be equivalent to those avail-able in urban settings,50 some rural HCPs lament their lack of palliative skills, training and mentoring.51,52 This insuffi-ciency is especially significant in the hospital setting as end-of-life services need to be integrated into rural hospitals53 as they often act as substitute inpatient hospices.54,55
It was frequently reported in the literature that effec-tive communication between the patient/family caregiver and HCPs and within multidisciplinary teams is essential but often lacking.9,56 Patients and family caregivers expressed poor communication as a barrier to receiving quality end-of-life care especially when care was frag-mented due to distance and the need to commute between different locations and HCPs. Few studies in the literature reported on information and communication technologies for end-of-life care57 and while mobile phones and the Internet21,25,32 enabled some participants to stay con-nected, no rural studies were identified reporting on the use of communication technology in rural settings or if terminally ill patients would accept this as an alternative to physical contact.3,26,35
An additional challenge faced by rural communities is the ageing population4,58 with many rural elderly living on their own20 as families disperse. While the elderly are less likely to complain and demand little, they bear significant stress.20 Participants in this review were reliant on support from neighbours and the community with informal net-works being an untapped resource59 and the focus for future research. Rural communities are endowed with an ‘incredible volunteer base’ and many ‘very generous
people … who are really genuinely concerned about the community’.48
In contrast to the literature which places a high priority on pain and symptom control,50,53 very few participants in developed countries indicated that this was a high priority. This is possibly due to the aims and specific focus of the included studies, and with most participants enrolled in palliative/hospice care, it is also possible these services were delivering satisfactory symptom management.56 By comparison, in Kenya and Malawi, where palliative care is becoming more integrated into mainstream health services,43 pain management remains a high priority. Effective palliative/hospice care also increases the chance of patients receiving care and dying in their preferred place, usually home,28,53 however, with the exception of Indigenous participants,18,30 and consistent with previous studies,53 many participants in this review were accepting of the transition to the community hospital.11,27,33
The results of this review show that while researchers continue to seek out the end-of-life care experiences and perspectives of rural patients and family caregivers, this research has mainly occurred in North America and mainly with elderly, Caucasian and cancer patients. More work is required on all aspects of end-of-life care from patient and family caregiver perspectives, especially those living in remote areas or indigenous communities within developed nations where formal HCP support is either very limited or absent; the experiences of patients with advanced chronic non-cancer disease; paediatric patients and their family caregivers; from other developing countries in Asia, the Pacific and South America; patients and family caregivers who have to relocate; the evaluation of technology in enhancing information and communication; and longitudi-nal studies examining changes in experiences and perspec-tives over time.
This review affirms that recruiting patients and their family caregivers at end-of-life is difficult,7,60 and even more so in rural regions, resulting in small sample sizes. However, it cannot be assumed that the findings from small sample studies are irrelevant and that the rural voice can be added in unison with urbanites. The risk here is that universal models of care may develop and ignore the spe-cific attributes of the rural psyche (‘stoicism, fatalism … self- reliance and rugged independence, coupled with a lessened sense of confidentiality and increased pressure to conform due to the smaller, more intimate nature of smaller rural environments’41) or the specific rural challenges such as distance and isolation. Collaborative, multi-site or multi-national research may address this issue.
Limitations
This systematic literature review had limitations in rela-tion to the quality of the identified studies and complete-ness of the review. Studies where titles, abstracts or
910 Palliative Medicine 31(10)
keywords omitted the selection criteria wording may have been overlooked; however, electronic searching was aug-mented by handsearching journals and reference lists. Some excluded studies included rural participants; how-ever, rural data were not identified, and so, this exclusion meant that some potentially informative perspectives were not gained. Generalisability of results is limited by small sample size; heterogeneity and inconsistency in rural/remote definition meaning studies reported on a variety of rural locations, populations and proximity to health ser-vices; recruitment bias (single-service recruitment with most participants receiving formal end-of-life care); pre-dominance of elderly participants who may be more accepting of their circumstances than young patients; high proportion of cancer patients; and North American bias.
Conclusion
It is necessary to explore end-of-life care experiences and perspectives of rural patients and their family caregivers, as valuable insights will be lost and rural patients/family car-egivers’ care may be compromised if their voices and needs are ignored. Common themes such as hardship of distance reduced access to palliative care, community support and importance of home and ‘country’ highlight the influence of rural location on end-of-life care. While the number of studies has increased since 2009,8 especially in North America, there still remains limited published rural studies reporting on patient and family caregivers’ experiences and perspectives on rural end-of-life care and further research is encouraged. The development of national and international collaborative work using a universal definition of ‘rural’ may begin to more clearly articulate the ‘rural’ voice.
Acknowledgements
S.R. was responsible for conception; design; acquisition, analysis and interpretation of data; drafting, revising and final manuscript. All co-authors contributed to review of titles and/or abstracts/full texts; analysis, quality assessment and interpretation of data; revising the article critically for important intellectual content; and approval of the near final version to be published.
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship and/or publication of this article.
References
1. World Health Organization. Palliative Care, Fact sheet no. 402 July 2015, http://www.who.int/mediacentre/factsheets/fs402/en/ (accessed 22 May 2016).
2. Cottrell L and Duggleby W. The ‘good death’: an integra-tive literature review. Palliat Support Care 2016; 16(6): 686–712.
3. Bakitas M, Elk R, Astin M, Ceronsky L, et al. Systematic review of palliative care in the rural setting. Cancer Control 2015; 22(4): 450–464.
4. Wilson DM, Fillion L, Thomas R, et al. The ‘good’ rural death: a report of an ethnographic study in Alberta, Canada. J Palliat Care 2009; 25(1): 21–29.
5. Wilson DM, Fillion L, Thomas R, et al. Planning and pro-viding for a good death using rural French-Canadian and English-Canadian insights. Rev Neurosci 2009; 20: 313–319.
6. Hart LG, Larson EH and Lishner DM. Rural definitions for health policy and research. Critical concepts for reaching pop-ulations at risk. Am J Public Health 2005; 95(7): 1149–1155.
7. Rainsford S, Bullen T and Rosenberg JP. Challenges of recruiting hospice patients with advanced cancer to research: reflections on a delirium screening study. Prog Palliat Care 2015; 23(1): 24–28.
8. Robinson CA, Pesut B, Bottorff JL, et al. Rural palliative care: a comprehensive review. J Palliat Med 2009; 12(3): 253–258.
9. Hughes PM, Ingleton MC and Clark D. Providing cancer and palliative care in rural areas: a review of patient and carer needs. J Palliat Care 2004; 20(1): 44–49.
10. Shamseer L, Moher D, Clarke M, et al. PRISMA-P Group. Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) 2015: elaboration and explanation. BMJ 2015; 349: g7647
11. Payne S, Hawker S and Kerr C. Experiences of end-of-life care in community hospitals. Health Soc Care Community 2007; 15(5): 494–501 (8 pp.).
12. Gomes B, Calanzani N, Gysels M, et al. Heterogeneity and changes in preferences for dying at home: a systematic review. BMC Palliat Care 2013; 12: 7.
13. Hannes K. Chapter 4: critical appraisal of qualitative research. In: Noyes J, Booth A, Hannes K, et al. (eds), Supplementary guidance for inclusion of qualitative research in cochrane systematic reviews of interventions. Version 1 (updated August 2011). Cochrane collaboration qualitative methods group. 2011, http://cqrmg.cochrane.org/supplemental-handbook-guidance (cited 23/1/16).
14. Brazil K, Kaasalainen S, Williams A, et al. A comparison of support needs between rural and urban family caregiv-ers providing palliative care. Am J Hosp Palliat Care 2014; 31(1): 13–19.
15. Brazil K, Kaasalainen S, Williams A, et al. Comparing the experiences of rural and urban family caregivers of the ter-minally ill. Rural Remote Health 2013; 13(1): 1–9.
16. Dale MJ and Johnston B. An exploration of the concerns of patients with inoperable lung cancer. Int J Palliat Nurs 2011; 17(6): 285–290.
17. Darer JD, Clarke DN, Sees AC, et al. Assessing the quality of death and dying in an integrated health care system in rural Pennsylvania. J Pain Symptom Manage 2015; 50(3): 343.e6–349.e6.
18. Dembinsky M. Exploring Yamatji perceptions and use of palliative care: an ethnographic study. Int J Palliat Nurs 2014; 20(8): 387–393.
Rainsford et al. 911
19. Devik SA, Enmarker I, Wiik GB, et al. Meanings of being old, living on one’s own and suffering from incurable cancer in rural Norway. Eur J Oncol Nurs 2013; 17(6): 781–787.
20. Wiik GB. Don’t become a burden and don’t complain: a case study of older persons suffering from incurable cancer and living alone in rural areas. Nurs Reports 2011; 1(e3): 7–14.
21. Devik SA, Hellzen O and Enmarker I. ‘Picking up the pieces’ – meanings of receiving home nursing care when being old and living with advanced cancer in a rural area. Int J Qual Stud Health Well-being 2015; 10: 28382.
22. Duggleby WD, Penz K, Leipert BD, et al. ‘I am part of the community but’. The changing context of rural living for persons with advanced cancer and their families. Rural Remote Health 2011; 11(3): 1733.
23. Duggleby WD, Penz KL, Goodridge DM, et al. The transi-tion experience of rural older persons with advanced cancer and their families: a grounded theory study. BMC Palliat Care 2010; 9: 1–5.
24. Duggleby WD, Williams A and Holstlander L. Hope of rural women caregivers of persons with advanced cancer: guilt, self-efficacy and mental health. Rural Remote Health 2014; 14: 2561.
25. Grant L, Brown J, Leng M, et al. Palliative care making a difference in rural Uganda, Kenya and Malawi: three rapid evaluation field studies. BMC Palliat Care 2011; 10: 8.
26. Hansen L, Cartwright JC and Craig CE. End-of-life care for rural-dwelling older adults and their primary family car-egivers. Res Gerontol Nurs 2012; 5(1): 6–15.
27. Hatcher I, Harms L, Walker B, et al. Rural palliative care transitions from home to hospital: carers’ experiences. Aust J Rural Health 2014; 22(4): 160–164.
28. Herce M, Elmore SN, Kalanga N, et al. Assessing and responding to palliative care needs in rural sub-Saharan Africa: results from a model intervention and situation anal-ysis in Malawi. PLoS ONE 2014; 9(10): 17.
29. Johnston BM, Milligan S, Foster C, et al. Self-care and end of life care – patients’ and carers’ experience a qualitative study utilising serial triangulated interviews. Support Care Cancer 2012; 20(8): 1619–1627.
30. Kelly L, Linkewich B, Cromarty H, et al. Palliative care of first nations people a qualitative study of bereaved family members. Can Fam Physician 2009; 55(4): 394.e7–395.e7.
31. Lockie SJ, Bottorff JL, Robinson CA, et al. Experiences of rural family caregivers who assist with commuting for pal-liative care. Can J Nurs Res. 2010; 42(1): 74–91.
32. Pesut B, Robinson CA, Bottorff JL, et al. On the road again: patient perspectives on commuting for palliative care. Palliat Support Care 2010; 8(2): 187–195.
33. Mixer S, Fornehed ML, Varney J, et al. Culturally congru-ent end-of-life care for rural Appalachian people and their families. J Hosp Palliat Nurs 2014; 16(8): 526–535.
34. Ostertag SG and Forman WB. End-of-life care in Hancock County, Maine: a community snapshot. Am J Hosp Palliat Med 2008; 25(2): 132–138 (7 pp.).
35. Pesut B, Bottorff JL and Robinson CA. Be known, be available, be mutual: a qualitative ethical analysis of social
values in rural palliative care. BMC Med Ethics 2011; 12: 19.
36. Pesut B, Robinson CA and Bottorff JL. Among neighbors: an ethnographic account of responsibilities in rural pallia-tive care. Palliat Support Care 2014; 12(2): 127–138.
37. Revier SS, Meiers SJ and Herth KA. The lived experience of hope in family caregivers caring for a terminally ill loved one. J Hosp Palliat Nurs 2012; 14(6): 438–446.
38. Tamannai M, Kaah J and Mbah G. An evaluation of a pallia-tive care outreach programme for children with Burkitt lym-phoma in rural Cameroon. Int J Palliat Nurs 2015; 21(7): 331–337 (7 pp.).
39. Williams A, Duggleby W, Eby J, et al. Hope against hope: exploring the hopes and challenges of rural female caregiv-ers of persons with advanced cancer. BMC Palliat Care 2013; 12(1): 44.
40. Moher D, Liberati A, Tetzlaff J, et al. Preferred report-ing items for systematic reviews and meta-analyses: the PRISMA statement. Ann Intern Med 2009; 151(4): 264–270.
41. Gessert C, Waring S, Bailey–Davis L, et al. Rural definition of health: a systematic literature review. BMC Public Health 2015; 15: 378.
42. Kirby S, Barlow V, Saurman E, et al. Are rural and remote patients, families and caregivers needs in life-limiting ill-ness different from those of urban dwellers? A narrative synthesis of the literature. Aust J Rural Health 2016; 24(5): 285–345.
43. Lynch T, Connor S and Clark D. Mapping levels of pal-liative care development: a global update. J Pain Symptom Manag 2013; 45(6): 1094–1106.
44. Ciemins EL, Brant J, Kersten D, et al. A quality analysis of patients and family perspectives of palliative care. J Palliat Med 2015; 18(3): 282–285.
45. Sandsdalen T, Hov R, Hoye S, et al. Patients’ preferences in palliative care: a systematic mixed studies review. Palliat Med 2015; 29(5): 399–419.
46. Steinhauser KE, Christakis NA, Clipp EC, et al. Factors considered important at the end of life by patients, family, physicians and other care providers. JAMA 2000; 284(19): 2476–2482.
47. Scott G, Whyler N and Grant GA. Study of family carers of people with a life-threatening illness: the carer’s need analy-sis. Int J Palliat Nurs 2001; 7(6): 290–297.
48. Kaasalainen S, Brazil K, Williams A, et al. Nurses’ experi-ences providing palliative care to individual living in rural communities: aspects of the physical residential setting. Rural Remote Heath 2014; 14: 2728.
49. Castledean H, Crooks V, Schuurman N, et al. ‘It’s not necessarily the distance on the map …’: using place as an analytic tool to elucidate geographical issues central to rural palliative care. Health and Place 2010; 16: 284–290.
50. Goodridge D and Duggleby W. Using a quality framework to assess rural palliative care. J Palliat Care 2010; 26(3): 141–150.
51. Fink RM, Oman KS, Youngwerth J, et al. A palliative care needs assessment of rural hospitals. J Palliat Med 2013; 16(6): 638–644.
912 Palliative Medicine 31(10)
52. Spice R, Paul LR and Biondo PD. Development of a rural palliative care program in the Calgary Zone of Alberta Health Services. J Pain Symptom Manage 2012; 43(5): 911–924.
53. Wilson DM, Justice C, Sheps S, et al. Planning and provid-ing end-of-life care in rural areas. National Rural Health Assoc 2006; 22(2): 174–181.
54. Payne S, Kerr C, Hawker S, et al. Community hospitals: an under-recognised resource for palliative care. J R Soc Med 2004; 97: 428–431.
55. Rainsford S, MacLeod RD and Glasgow NJ. Place of death in rural palliative care: a systematic review. Palliat Med 2016; 30(8): 745–763.
56. Ventura AD, Burney S, Brooker J, et al. Home-based pallia-tive care: a systematic literature review of the self reported unmet needs of patients and carers. Palliat Med 2014; 28(5): 391–402.
57. Ostherr K, Killoran P, Shegog R, et al. Death in the digital age: a systematic review of information and communication technologies in end-of-life care. J Palliat Med 2016; 19(4): 408–420.
58. Hugo G, Feist H, Tan G and Harris K. Population dynamics in regional Australia. Regional Australia Institute 2015. http://www.regionalaustralia.org.au/wp-content/uploads/ 2015/01/FINAL-Population-Dynamics-in-Regional-Australia. pdf (accessed 23 December 2016).
59. Horsfall D, Leonard R, Noonan K, et al. Working together-apart: exploring the relationships between formal and infor-mal care networks for people dying at home. Prog Palliat Care 2013; 21(6): 331–336.
60. Wohleber AM, McKitrick DS and Davis SE. Designing research with hospice and palliative care populations. Am J Hosp Palliat Med 2012; 29(5): 335–345.
Recommended